State-Dependent Oral CRISPR Delivery: A Systems Framework for Uptake, EBV Targeting, and Gene Editing Efficiency

A CYNAERA framework for modeling environmental burden, viral persistence, and flare dynamics in oral CRISPR-based EBV therapies

This paper is part of the growing CYNAERA Remission™ Library redefining how oral delivery gene editing is applied to chronic, immune-volatile conditions through state-dependent intervention.

By Cynthia Adinig

1. Executive Summary

CRISPR-based therapies are typically evaluated based on targeting precision and molecular design. However, one of the most critical determinants of success remains under-modeled: delivery efficiency within dynamic biological systems. This gap is especially pronounced in oral CRISPR approaches, where gastrointestinal variability, immune activation, and environmental burden directly influence uptake, distribution, and therapeutic engagement.

This paper introduces a state-dependent oral CRISPR delivery framework, positioning delivery not as a static engineering constraint, but as a biologically variable process shaped by system load, regulatory capacity, and environmental exposure. Unlike CRISPR Remission™, which focuses on pathway targeting and stabilization, this framework focuses specifically on how gene editing interventions are successfully delivered, absorbed, and tolerated in real-world physiologic conditions.

This distinction is particularly relevant for Epstein–Barr virus (EBV)-associated disease modeling and persistent infection states, where viral latency, immune signaling, and mucosal immune interaction create highly variable conditions for intervention. Oral CRISPR delivery introduces additional complexity, as therapeutic material must navigate gastrointestinal barriers, microbiome interaction, and immune surveillance before reaching target tissues. These factors are not peripheral. They are central determinants of whether gene editing succeeds or fails.

Current CRISPR literature largely treats delivery as a vector optimization problem, focusing on viral vectors, lipid nanoparticles, or targeting specificity (Naldini, 2015; Pardi et al., 2018). While these approaches are necessary, they are incomplete. They do not account for host-state variability, including inflammatory load, autonomic instability, and environmental burden, all of which influence delivery efficiency and cellular uptake. This paper reframes CRISPR delivery as a state-dependent system interaction, where timing, physiologic readiness, and environmental context determine whether therapeutic input is successfully integrated.

Building on CYNAERA models, including the Input Load Collapse Model (ILCM) and STAIR stabilization architecture, we demonstrate that apparent delivery failure is frequently not a function of molecular inefficiency, but of system-level misalignment between therapeutic input and host state. This framework directly informs the CRISPR Remission™ architecture, a flare-aware gene editing pathway model designed for immune-volatile chronic disease, which has been accepted for presentation at the CRISPR Medicine Conference (CRISPRMED26). The CRISPR Remission™ approach operationalizes state-dependent delivery by integrating biologic timing, environmental load, and system readiness into gene-editing strategy.

We further define a state-dependent safety architecture for immune-volatile populations, showing that adverse events are often threshold-dependent and systematically misclassified under current clinical models. By integrating graded exposure protocols, environmental load modeling, and system-state gating, this framework enables safer and more interpretable therapeutic deployment across highly reactive patient populations.

This approach reframes variability from an irreducible property of complex disease into a measurable and controllable parameter. In clinical trials, state-dependent alignment has the potential to reduce dropout, improve signal clarity, and prevent premature dismissal of viable therapies. At scale, these effects translate into meaningful economic impact by reducing inefficiency while expanding the treatable population across severity levels and age groups. By positioning CRISPR delivery, particularly oral CRISPR delivery, as a dynamic systems problem rather than a purely molecular one, this framework provides a pathway for improving therapeutic consistency, redefining safety evaluation, and increasing the clinical and economic viability of gene-based interventions in complex chronic disease populations.

2. Static Delivery Models in Dynamic Host–Virus Systems 

CRISPR-based approaches to EBV and other persistent viral infections have increasingly focused on identifying and disrupting host factors that enable viral persistence and proliferation. Genome-wide CRISPR-Cas9 screens in EBV-infected B cells have demonstrated that viral latency programs, particularly latency III, establish dependency on specific host regulatory pathways, including NF-κB signaling and downstream transcriptional networks such as the LMP1–JunB axis [Wang et al., 2019; Kieser & Sterz, 2015; Ma et al., 2017]. These pathways actively repress cell cycle inhibitors such as CDKN2C/p18INK4c, enabling continued proliferation of infected cells. As a result, EBV-driven disease is governed not by a single viral target, but by a host–virus regulatory system in which viral proteins and host transcription factors interact to sustain pathological states.

Despite this complexity, delivery models for CRISPR therapeutics remain largely static. Success is typically evaluated in terms of vector efficiency, target accessibility, and editing precision, while host system state is treated as a secondary or uncontrolled variable [Lino et al., 2018; Mitchell et al., 2021]. This introduces a fundamental mismatch between therapeutic design and biological reality. In EBV-associated conditions and broader IACC states, host regulatory networks are continuously modulated by viral signaling. LMP1-driven NF-κB activation, MAPK signaling, and cytokine modulation produce a system characterized by persistent activation, feedback coupling, and instability [Young et al., 2016; Eliopoulos & Young, 2001]. Under these conditions, the effectiveness of a CRISPR intervention cannot be separated from the state of the system at the time of delivery.

For example, a CRISPR strategy targeting JunB or related transcriptional pathways may theoretically disrupt EBV-driven proliferation. However, if introduced during a period of heightened immune activation or systemic instability, the intervention may encounter altered transcriptional accessibility, increased inflammatory signaling, or reduced cellular receptivity. These factors can attenuate editing efficiency or amplify unintended responses, consistent with observed variability in gene and cell therapy outcomes across heterogeneous host states [Naldini, 2015; June et al., 2018].

Within current models, these inconsistencies are typically attributed to technical limitations in delivery systems or guide design. However, this interpretation overlooks a primary driver of variability: unmeasured biological state at the time of intervention. As a result, delivery outcomes that are fundamentally state-dependent are misclassified as stochastic or technical failure, limiting the ability to optimize therapeutic strategies and obscuring the true determinants of CRISPR efficacy.

3. Latency-Driven Instability and State-Dependent Outcomes 

EBV latency programs provide a clear model of how persistent infection reshapes host biology into a dynamic and continuously modulated system. During latency III, viral proteins such as LMP1 mimic constitutive receptor signaling, activating NF-κB, MAPK, and JAK/STAT pathways while altering host transcriptional regulation [Kieser & Sterz, 2015; Young et al., 2016].

These effects are required to maintain the proliferative state of infected B cells and are sustained through ongoing viral–host interaction rather than transient signaling events.

CRISPR screening data further demonstrate that host factors such as JunB are not merely associated with this state, but are required for its maintenance. JunB expression is actively sustained by LMP1 signaling and contributes to repression of cell cycle inhibitors that would otherwise halt proliferation [Wang et al., 2019]. Disruption of this axis alters cell cycle progression and reduces proliferative capacity, confirming that EBV-driven disease is sustained through tightly coupled host–virus regulatory interactions.

Within this architecture, the host system does not operate around a stable baseline. Instead, it fluctuates in response to internal and external inputs, including immune activation, environmental exposure, autonomic variability, and cumulative physiological load. These fluctuations alter transcriptional activity, intracellular signaling environments, and metabolic conditions that directly influence cellular response to gene-editing interventions. As a result, CRISPR delivery outcomes become state-dependent. During periods of elevated system instability, characterized by heightened inflammatory signaling and active pathway engagement, CRISPR components may encounter altered intracellular processing, reduced editing efficiency, or amplified immune responses. These conditions increase the likelihood of incomplete editing, off-target effects, or adverse reactions. This behavior is consistent with systems-level models in which cumulative input exceeds regulatory capacity, resulting in amplified responses and reduced tolerance to additional intervention [Medzhitov, 2008; McEwen, 2007; Sterling, 2012].

Conversely, during periods of relative stability, when inflammatory signaling and system volatility are reduced, the same interventions may achieve more consistent delivery, improved editing efficiency, and greater therapeutic impact. This contrast demonstrates that delivery success is not solely a function of vector design or targeting precision. It is determined by alignment between intervention and system state. Recognizing EBV-driven disease as a dynamic host–virus system reframes CRISPR delivery as a problem of timing, context, and biologic readiness. This perspective is consistent with systems biology frameworks demonstrating that cellular responsiveness to intervention is conditioned by network state and feedback structure rather than isolated pathway activation [Kitano, 2002; Barabási et al., 2011].

Advances in CRISPR-based therapeutics have demonstrated that precise targeting of disease-relevant pathways is biologically achievable. However, the ability to identify and target critical nodes within a host–virus regulatory network does not ensure consistent therapeutic success. The limiting factor is not precision at the level of targeting, but alignment at the level of system state. CRISPR delivery must therefore be understood not as a discrete molecular event, but as an interaction between an engineered intervention and a dynamic biological system in which outcomes are co-determined by intervention properties and system conditions at the time of delivery.

4. From Molecular Targeting to State-Dependent Delivery Architecture 

Advances in CRISPR-based therapeutics have demonstrated that precise targeting of disease-relevant pathways is biologically achievable. In EBV-associated systems, interventions directed at host dependency factors such as JunB or downstream cell cycle regulators have the potential to disrupt viral-driven proliferation and alter disease trajectories [Wang et al., 2019; Ma et al., 2017]. However, the ability to identify and target critical nodes within a host–virus regulatory network does not ensure consistent therapeutic success. The limiting factor is not precision at the level of targeting, but alignment at the level of system state.

Current delivery paradigms treat CRISPR intervention as a discrete event governed by molecular rules of uptake, editing, and repair. This model assumes that the biological environment is sufficiently stable to allow predictable interaction between therapeutic systems and their targets. In immune-volatile conditions, this assumption fails. CRISPR delivery must instead be understood as an interaction between an engineered intervention and a dynamic biological network, in which outcomes are co-determined by intervention properties and system state at the time of delivery. This perspective aligns with systems biology frameworks demonstrating that cellular responsiveness to intervention is conditioned by network state and feedback structure rather than isolated pathway activation [Kitano, 2002; Barabási et al., 2011].

Within this context, delivery failure is frequently not a consequence of inadequate vector design or insufficient targeting specificity. It is the result of introducing therapeutic input into a system operating outside its capacity to integrate that input. This distinction is critical. It reframes delivery from a problem of transport and targeting to a problem of timing, threshold, and biologic readiness. To formalize this relationship, system state must be represented as a function of interacting biological domains that collectively determine how input is processed. This transition from molecular targeting to state-dependent delivery architecture, provides the foundation for a more predictive and controllable model of CRISPR intervention.

4.1 Defining System State in Immune-Volatile Conditions  

System state can be conceptualized as a multidimensional representation of biological activity across interacting domains:

• N(t): Neural and nociceptive signaling

• I(t): Inflammatory and immune signaling load

• A(t): Autonomic regulation and vascular dynamics

• C(t): Central processing and sensitization

• E(t): Environmental and external exposure load

Each variable represents a time-dependent function capturing both magnitude and variability. These domains are not independent. They are coupled through feedback mechanisms that allow perturbation in one domain to propagate across the system. The overall system state can be expressed as:

S(t) = w₁N(t) + w₂I(t) + w₃A(t) + w₄C(t) + w₅E(t)

where weights reflect phenotype-specific contributions.

This representation is consistent with multi-system disease models observed in post-infectious

syndromes, dysautonomia, and neuroimmune conditions, where interactions between immune, autonomic, and environmental domains produce emergent behavior not predictable from any single subsystem [VanElzakker et al., 2019; Vernon et al., 2010; Klimas et al., 2012].

4.2 Instability, Load, and the Threshold for Intervention  

System behavior in immune-volatile conditions is non-linear. Cumulative input across domains can exceed regulatory capacity, producing amplified responses and reduced tolerance to additional stimuli. This can be expressed as:

dS/dt ∝ (N + I + A + C + E) − R

where R represents regulatory capacity.

When cumulative input exceeds R, the system transitions into a high-instability state characterized by:

• amplified inflammatory signaling

• increased autonomic variability

• reduced tolerance to intervention

• nonlinear response to additional inputs

This behavior is consistent with models of allostatic load and stress-induced system dysregulation, in which cumulative physiological burden reduces adaptive capacity and increases sensitivity to additional stressors [McEwen & Wingfield, 2003; Sterling, 2012]. Within this state, CRISPR delivery is not neutral. It represents an additional input into an already constrained system. Even with precise targeting, the system may respond with reduced uptake efficiency, altered intracellular processing, or amplified immune activation. These outcomes are frequently interpreted as delivery failure. Within a state-dependent framework, they represent mismatch between intervention and system readiness.

4.3 Stabilization Windows and Timing of CRISPR Delivery  

If instability reduces the probability of successful intervention, periods of reduced volatility represent optimal windows for delivery. Stabilization windows can be defined as intervals in which both system load and short-term variability decrease across domains.

Let σₙ, σᵢ, σₐ, σ𝚌, and σₑ represent variability in each domain. A stabilization function can be defined as:

W(t) ∝ 1 / (ασₙ + βσᵢ + γσₐ + δσ𝚌 + εσₑ)

A stabilization window exists when:

W(t) > θ and dW/dt < 0

These conditions identify periods in which:

• system volatility is reduced

• regulatory processes are dominant

• tolerance capacity is effectively expanded

During these windows, CRISPR delivery is more likely to achieve consistent uptake, predictable pathway interaction, and reduced immune reactivity. Temporal sensitivity to intervention has been observed across multiple biological systems, including circadian-regulated immune response, metabolic timing, and pharmacodynamic variability, reinforcing that timing is a primary determinant of therapeutic outcome rather than a secondary consideration [Cermakian et al., 2013; Dallmann et al., 2014].

4.4 Implications for CRISPR Targeting in EBV Systems  

In EBV-driven systems, CRISPR strategies targeting transcriptional regulators such as JunB operate within pathways that are dynamically modulated by viral signaling. During periods of elevated NF-κB activation and immune signaling, transcriptional networks may exhibit increased resistance to perturbation or altered responsiveness. Conversely, during periods of reduced signaling intensity, the same interventions may produce more effective pathway modulation.

This demonstrates that targeting specificity and delivery timing are interdependent variables.

Failure to account for this interaction can lead to underestimation of therapeutic potential, as interventions that are effective under optimal conditions appear inconsistent when applied across heterogeneous system states.

4.5 Transition to Delivery-Aware Therapeutic Design  

Incorporating system state into CRISPR delivery design represents a transition from static to adaptive therapeutic architecture.

Within this model:

• delivery is aligned with system readiness

• variability is treated as a measurable parameter

• intervention timing becomes a controllable variable

This shift has implications across therapeutic development, clinical trial design, and patient-level application. By embedding system-state awareness into delivery frameworks, CRISPR therapeutics can move from probabilistic outcomes toward predictable, reproducible intervention strategies in complex biological systems.

5. Oral CRISPR Delivery and Route-Dependent Tolerance  

Oral delivery of nucleic acid–based therapeutics represents a critical frontier in translational medicine due to its scalability, accessibility, and potential for long-term adherence. However, it introduces biological complexity that extends beyond traditional considerations of stability and absorption. Unlike intravenous delivery, oral delivery requires navigation of the gastrointestinal environment, immune surveillance systems, and variable absorption dynamics prior to systemic distribution [Kulkarni et al., 2021; Mitchell et al., 2021].In immune-volatile systems, these challenges are not solely technical. They are systemic.

5.1 Structural Barriers to Oral CRISPR Delivery 

Despite its theoretical advantages, oral delivery of nucleic acid therapeutics, including CRISPR-based systems, remains largely preclinical due to a convergence of biological and physicochemical barriers that limit stability, absorption, and systemic distribution.

1. Gastrointestinal Degradation Gastric acidity and digestive enzymes rapidly break down unprotected RNA and DNA constructs before they reach absorptive surfaces. [Pardi et al., 2018; Hou et al., 2021].

2. Mucus Penetration The intestinal mucus layer traps nanoparticle carriers and restricts diffusion toward the epithelial surface. Particle size, charge, and surface chemistry all affect penetration. [Ensign et al., 2012; Lai et al., 2009].

3. Epithelial Transport Tight junctions restrict paracellular transport, transcellular uptake remains inefficient, and endosomal escape further limits intracellular delivery efficiency. [Mitchell et al., 2021].

4. Systemic Uptake and Immune Activation Bioavailability after intestinal absorption remains low, limiting access to target tissues. At the same time, gut-associated immune responses can trigger clearance, inflammatory signaling, or intolerance. [Kulkarni et al., 2021].

In immune-volatile populations, including those with mast cell activation and post-infectious dysregulation, this response may be amplified, leading to rapid clearance, inflammatory signaling, or intolerance to otherwise viable formulations [Mowat & Agace, 2014; Afrin et al., 2016]. Taken together, these barriers explain why oral delivery of nucleic acid therapeutics, including mRNA and CRISPR systems, has not yet achieved consistent clinical translation. Existing approaches have focused primarily on overcoming individual barriers through formulation engineering. However, this strategy assumes that biological variability can be controlled at the level of the delivery vehicle alone.

Within a state-dependent framework, these barriers are not static. They are dynamically modulated by system conditions. Gastrointestinal permeability, mucus composition, immune activation, and absorption efficiency all vary in response to inflammatory signaling, autonomic regulation, environmental exposure, and cumulative system load. As a result, delivery success is not determined solely by formulation, but by the interaction between formulation and system state at the time of administration.

5.2 The Gastrointestinal System as a Dynamic Interface  

The gastrointestinal tract functions as an active immunological interface integrating signals from the microbiome, environmental exposures, and host immune activity. In conditions characterized by mast cell activation, dysautonomia, and post-infectious immune dysregulation, this interface becomes highly variable. Factors influencing oral delivery include:

• fluctuating gut permeability

• mast cell activation within GI tissue

• microbiome-dependent metabolic variability

• autonomic regulation of motility and perfusion

These dynamics are well-documented in disorders involving gut–brain and neuroimmune interaction, where gastrointestinal variability directly influences systemic immune behavior [Mayer et al., 2015; Camilleri, 2021; Afrin et al., 2016]. As a result, oral therapeutic success becomes dependent on system state rather than formulation alone.

5.3 Route-Dependent Variability in Immune-Volatile Systems  

Differences between oral and intravenous delivery highlight the impact of system state on therapeutic tolerance. Intravenous delivery bypasses gastrointestinal variability and delivers therapeutics directly into circulation. Oral delivery, by contrast, requires interaction with a system that may be actively dysregulated.

In immune-volatile conditions, the same therapeutic may be:

• tolerated intravenously

• partially effective orally under stable conditions

• poorly tolerated orally during periods of instability

This variability reflects differences in exposure timing, immune engagement, and cumulative system load. Within threshold-based models, oral delivery introduces additional input at the gastrointestinal level. When baseline load is elevated, this input may exceed tolerance thresholds, resulting in amplified responses that are often misclassified as intolerance. These responses frequently present without visible allergic markers, instead manifesting as cardiovascular, neurologic, or gastrointestinal symptoms, consistent with emerging recognition of non-classical immune activation patterns in mast cell–associated and post-infectious conditions [Weinstock et al., 2021].

5.4 Environmental Modulation of Oral CRISPR Delivery 

Environmental exposure functions as an active input into system state rather than an external modifier. Variables such as particulate matter, ozone, mold burden, humidity, and barometric pressure shifts influence inflammatory signaling, autonomic regulation, and gastrointestinal permeability [Kampa & Castanas, 2008; Mendell et al., 2011]. These inputs contribute directly to cumulative system load:

L(t) = I(t) + A(t) + G(t) + M(t) + E(t)

Within this framework, oral CRISPR delivery is co-determined by environmental conditions at the time of intervention. A therapy tolerated under low environmental load may become destabilizing under elevated exposure conditions due to increased baseline system load rather than changes in the therapeutic itself. This effect is consistent with evidence demonstrating that environmental exposures can suppress treatment response and increase symptom burden in chronic illness populations [Dales et al., 2008; Breysse et al., 2010].

5.5 Environmental Load Modeling and Threshold-Dependent Delivery  

Oral delivery can be modeled as a threshold-dependent process:

Successful delivery occurs when:

L(t) < T

Environmental inputs increase E(t), raising total system load and reducing tolerance margin.

This relationship can be operationalized using predictive environmental modeling frameworks such as CYNAERA’s VitalGuard™, which integrates exposure metrics, condition-specific sensitivities, and regional factors into a composite load estimate. As environmental load increases:

L(t) = L_base + E(t)

When:

L_base + E(t) ≥ T

delivery outcomes shift toward reduced absorption, increased immune reactivity, or intolerance.

This explains observed variability in treatment tolerance across days with differing environmental conditions.

5.6 Integration with State-Dependent Delivery Architecture  

Effective oral CRISPR delivery requires integration into a broader state-dependent framework.

This includes:

• assessment of system state prior to intervention

• alignment with stabilization windows

• route selection based on tolerance probability

• adaptive dosing strategies

These principles align with stabilization-first architectures such as STAIR, in which intervention follows reduction of inflammatory, autonomic, and environmental load. Within this model, oral delivery becomes conditional rather than universal, improving both tolerability and efficacy by aligning intervention with system readiness.

6. From Targeting to Remission: A State-Dependent Intervention Framework

CRISPR-based therapeutics have established that disease-driving pathways in persistent viral and immune-mediated conditions can be precisely identified and modified [Doudna & Charpentier, 2014; Hsu et al., 2014]. Yet precision at the level of targeting does not reliably translate into remission in immune-volatile systems. The central limitation is not simply whether an edit can be made, but whether the system receiving that edit is in a state capable of integrating it.

In these conditions, therapeutic response is governed by the interaction between molecular targeting, delivery route, and biologic state at the time of intervention. System behavior is shaped by multiple domains operating simultaneously, including inflammatory activity, autonomic regulation, mast cell reactivity, gastrointestinal stability, environmental exposure, hormonal fluctuation, and viral persistence. A useful representation of this relationship is:

S(t) = w₁I(t) + w₂A(t) + w₃M(t) + w₄G(t) + w₅E(t) + w₆H(t) + w₇V(t)

where system state reflects the weighted contribution of interacting biological domains rather than any single mechanism in isolation. This type of multi-domain representation is consistent with systems biology and network medicine frameworks that model disease behavior as the product of interacting rather than isolated variables [Kitano, 2002; Barabási et al., 2011].

From this, cumulative system load can be expressed as:

L(t) = I + A + M + G + E + H + V

As total load approaches or exceeds the system’s tolerance threshold, T, response becomes increasingly nonlinear. Under these conditions, even a well-designed intervention may produce inconsistent uptake, altered intracellular processing, amplified immune responses, or incomplete therapeutic effect. This relationship can be summarized as:

Reaction Risk ∝ L(t) / T

What is often interpreted as delivery failure or therapeutic intolerance may therefore reflect system-state misalignment rather than failure of molecular design. In other words, the same intervention can appear ineffective in one biologic context and effective in another, not because the target has changed, but because the system’s capacity to process intervention has changed. This threshold-dependent logic is consistent with models of allostatic load and adaptive regulation, in which cumulative burden reduces the ability of biologic systems to absorb additional input without destabilization [McEwen, 2007; Sterling, 2012].

This makes stabilization a precondition for remission-oriented intervention rather than an optional supportive step. A viable treatment window emerges when short-term variability across the major domains declines and regulatory capacity begins to exceed cumulative load. This can be represented through a stabilization function:

W(t) ∝ 1 / (ασᵢ + βσₐ + γσₘ + δσg + εσₑ + ζσₕ + ησᵥ)

A meaningful intervention window exists when:

W(t) > θ and dW/dt < 0

These conditions identify periods in which the system is not only less volatile, but actively moving away from destabilization. That distinction matters. A low-load moment that is immediately preceding a flare is not equivalent to a low-load moment within a sustained trajectory of stabilization. Therapeutic success depends not only on static load level, but on the direction of system movement over time. This emphasis on temporal system position is consistent with broader biologic literature showing that response to intervention depends on dynamic state rather than static baseline alone [Cermakian et al., 2013; Kitano, 2002].

Within this framework, STAIR functions as a pre-conditioning architecture that shifts the system toward a more intervention-ready state. By reducing inflammatory burden, improving autonomic regulation, minimizing environmental stressors, and supporting gastrointestinal stability, STAIR lowers L(t) while functionally expanding the system’s tolerance threshold. This does not guarantee remission on its own, but it changes the conditions under which remission becomes biologically achievable. This logic is also consistent with CYNAERA’s state-dependent intervention models, which treat readiness as a measurable prerequisite for durable response rather than as a secondary supportive factor.

Once a stabilization window is identified, intervention should not be conceptualized as a single-point event. It should be treated as a staged process in which therapeutic input is introduced in a way the system can absorb without crossing threshold. The underlying sequence can be expressed as:

L₀ > L₁ → W(t) ↑ → Controlled Input → Adaptive Response

This progression reflects a shift from high baseline instability to lower load, followed by delivery during a period of increasing biologic readiness. Under these conditions, CRISPR intervention is more likely to engage target pathways consistently and less likely to function as an additional destabilizing input. Monitoring then becomes essential, not simply to detect overt adverse events, but to determine whether the system is integrating intervention, approaching overload, or requiring recalibration before escalation. This staged logic is consistent with graded exposure and desensitization principles used in highly reactive populations, where progressive introduction improves both safety and interpretability of response [Castells, 2009; Torres et al., 2019].

In this sense, remission should not be understood as the automatic consequence of editing a relevant target. It is better understood as the outcome of successful alignment between intervention and system capacity. A remission-compatible state exists when regulatory capacity exceeds cumulative load and therapeutic input can be introduced without overwhelming the system. Formally, this can be represented as:

R(t) > L(t) and U(t) + L(t) < T

When these conditions are met, intervention is more likely to be integrated rather than resisted, variability is reduced, and therapeutic effects have a greater chance of becoming durable. This reframes remission from a purely molecular endpoint into a systems-level achievement emerging from timing, threshold management, and biologic readiness. That framing is also aligned with CYNAERA’s remission-oriented architecture, including CRISPR Remission™, which applies flare-aware timing and system readiness to gene-editing strategy in immune-volatile disease contexts.

That is the central implication of a state-dependent intervention model. CRISPR does not fail simply because a target is insufficient or a vector is imperfect. In many cases, it fails because intervention is introduced into a system that is not prepared to receive it. Once that mismatch is addressed, remission becomes less a matter of chance and more a matter of engineering alignment between therapeutic design and system state.

7. State-Dependent Safety and Adverse Reaction Architecture

Safety in immune-volatile systems cannot be defined as a fixed property of a therapeutic. It is determined by the interaction between cumulative system load and tolerance capacity at the time of exposure. This relationship can be expressed as:

Reaction Probability ∝ L(t) / T

When baseline load is elevated, even low-level inputs may exceed threshold and trigger adverse responses. When load is reduced, the same inputs may be tolerated. This threshold-dependent behavior aligns with broader models of physiologic stress and immune reactivity, in which cumulative burden reduces adaptive capacity and increases sensitivity to additional inputs [McEwen, 2007; Sterling, 2012].

A major limitation of current safety frameworks is their reliance on visible indicators such as rash or swelling. In immune-volatile conditions, including mast cell activation syndromes and post-infectious dysregulation, clinically significant reactions frequently occur without cutaneous signs. Instead, they present through cardiovascular instability, neurologic symptoms, respiratory changes, or gastrointestinal distress [Afrin et al., 2016; Weinstock et al., 2021; Akin, 2017].

These reactions are often misclassified or dismissed, contributing to both patient harm and distortion of clinical safety data. This gap reflects a structural limitation in how adverse events are defined and detected rather than a lack of underlying physiologic response. To address this, intervention must be introduced in a way that respects threshold dynamics rather than assuming linear dose–response behavior.

7.3 Graded Exposure Model  

To prevent threshold exceedance, intervention should be introduced progressively rather than as a single full-intensity input. The graded exposure sequence can be modeled as:

Input₁ < Input₂ < Input₃ < Input₄

Each stage is separated by sufficient time to observe both immediate and delayed system response. This approach is consistent with desensitization protocols and ultra-low-dose introduction strategies used in allergy, mast cell disorders, and drug hypersensitivity management, where gradual exposure improves tolerance and reduces risk of systemic reaction [Castells, 2017; Torres et al., 2019].

Operationally, this progression includes:

• external exposure to assess immediate reactivity

• mucosal or localized exposure to test low-level tolerance

• micro-dose systemic introduction below threshold

• incremental escalation contingent on maintained stability

This sequence reduces abrupt load increases, improves interpretability of response, and allows adaptive accommodation without triggering system destabilization.

7.4 Route-Dependent Load Distribution  

Route of administration modifies how therapeutic input is distributed across the system. This can be expressed as:

L(t) = L_systemic + L_route-specific

Route-specific load varies based on delivery pathway. Oral delivery introduces input through gastrointestinal and mast cell–active pathways, increasing dependence on gut permeability, immune signaling, and autonomic regulation. Intravenous delivery bypasses gastrointestinal variability but contributes directly to systemic load, increasing reliance on global regulatory capacity. Localized delivery constrains distribution but may still produce regional or systemic effects through immune signaling and circulation. These differences are consistent with pharmacokinetic and immunologic studies demonstrating that route of administration alters both immune engagement and systemic response patterns [Mitchell et al., 2021; Pardi et al., 2018].

Environmental and physiological variables continuously shift both L(t) and T. Air quality, mold exposure, hormonal fluctuation, circadian disruption, and viral activity can alter inflammatory signaling and autonomic balance over short time intervals [Kampa & Castanas, 2008; Mendell et al., 2011; Cermakian et al., 2013]. Without accounting for these inputs, safety assessments remain incomplete.

7.5 Implications for Clinical Trials and Safety Interpretation  

In clinical trials, failure to incorporate state-dependent safety leads to systematic misclassification of adverse events, increased dropout among unstable participants, and underestimation of therapeutic efficacy. Data generated under these conditions reflects uncontrolled variability rather than true intervention performance.

A state-dependent safety architecture addresses this by integrating:

• system-state assessment prior to intervention

• graded exposure to prevent threshold exceedance

• context-aware interpretation of adverse events

• stabilization-based reintroduction rather than exclusion

Within this model, safety is defined not only by the absence of reaction, but by the ability to introduce intervention without exceeding system capacity. This approach aligns with CYNAERA’s broader system-state frameworks, including the Input Load Collapse Model and STAIR stabilization architecture, which demonstrate that adverse responses often reflect threshold dynamics rather than intrinsic therapeutic toxicity. The core principle remains consistent: absence of visible symptoms does not indicate absence of biological response, and presence of reaction does not necessarily indicate therapeutic failure. In immune-volatile systems, both safety and efficacy must be interpreted in relation to system state.

8. Translational Architecture and Deployment 

The transition from conceptual framework to clinical application is not a matter of implementation alone. It is a matter of correcting a structural mismatch between how therapies are currently deployed and how immune-volatile systems actually behave. In existing models, patient variability is treated as an unavoidable complication. In practice, much of that variability is the predictable result of introducing intervention into systems operating at different levels of stability. At a systems level, this variability reflects differences in how biological domains contribute to overall state. This can be expressed as:

S(t) = w₁I(t) + w₂A(t) + w₃M(t) + w₄G(t) + w₅E(t) + w₆H(t) + w₇V(t)

These weights determine how the system processes input. A patient with high inflammatory weighting operates closer to threshold than one with dominant autonomic variability but lower inflammatory burden. A patient with active viral persistence may experience continuous baseline elevation in I(t) and V(t), narrowing tolerance even before intervention is introduced [Proal & VanElzakker, 2021; Choutka et al., 2022].

This has direct implications for CRISPR therapeutics. Unlike small molecules, CRISPR systems are not simply interacting with receptors or pathways. They are entering cells, engaging transcriptional machinery, and altering regulatory networks that may already be in a state of active modulation [Doudna & Charpentier, 2014; Hsu et al., 2014]. In EBV-associated systems, for example, viral latency programs continuously reshape host transcriptional behavior. This means that the cellular environment into which CRISPR is delivered is not neutral. It is actively shifting [Young et al., 2016].

When interventions are applied without accounting for this, outcomes become inconsistent in ways that appear stochastic but are not. Editing efficiency may vary not because of guide RNA design, but because transcriptional accessibility and cellular state differ at the moment of delivery. Immune responses to delivery vectors may be amplified in high-load states. Adverse reactions may cluster in participants who enter trials at peak instability. This is why stratification is not simply useful. It is required. These dynamics are consistent with variability observed in gene therapy and immunotherapy across heterogeneous host states [Naldini, 2015; June et al., 2018].

By identifying dominant system drivers, intervention can be matched to the conditions under which it is most likely to succeed. This includes determining how much stabilization is required, which delivery route is appropriate, and when intervention should occur. Without this alignment, trial outcomes reflect uncontrolled variability rather than therapeutic potential.

Operational Trial Alignment

The consequences of ignoring system state are most visible in clinical trials. Current designs assume that variability can be averaged out across participants. In immune-volatile populations, this assumption introduces systematic error. Participants enter trials with different baseline loads. Interventions are applied at arbitrary points relative to system state. Adverse events are interpreted without context. The result is a dataset in which variability is attributed to heterogeneity rather than to timing and threshold dynamics. A state-dependent trial framework corrects this by structuring intervention around system readiness.

Phase 1: Stabilization (STAIR) Reducing baseline load shifts participants away from threshold, increasing the likelihood that intervention will be tolerated and measurable.

Phase 2: Readiness gating Intervention is initiated only when stabilization conditions are met, consistent with W(t) > θ, ensuring that participants are not in active destabilization.

Phase 3: Controlled introduction Therapeutic input is introduced below threshold, reducing the risk of immediate overload responses that obscure efficacy.

Phase 4: Context-aware monitoring Adverse events are interpreted relative to system state, allowing differentiation between true intolerance and state-dependent reaction.

Phase 5: Stabilization-based re-challenge Participants who react are not excluded outright but are stabilized and reintroduced, preserving signal rather than discarding it.

This structure does more than improve safety. It improves data quality. It reduces noise, clarifies response patterns, and allows therapeutic effects to be observed under conditions in which they are biologically possible. Delivery strategy must also be reconsidered in this context. It is not sufficient to optimize delivery vehicles while ignoring the system receiving them.

Effective Delivery = f(Route × Timing × System State)

This relationship is not symmetrical. Changes in system state can override improvements in delivery technology. Oral delivery, for example, depends heavily on gastrointestinal stability and mast cell activity. In high-load states, absorption may be impaired or immune responses amplified. Intravenous delivery bypasses part of this variability but increases reliance on systemic regulatory capacity, making it vulnerable to threshold effects at the whole-system level [Mitchell et al., 2021; Pardi et al., 2018].

The implication is that delivery must be selected dynamically. A route that performs well under stabilized conditions may fail under high load, not because of technical limitations, but because the system cannot process the input. In real-world settings, these dynamics are amplified. Environmental exposure, including air quality and mold, can shift inflammatory load within hours. Hormonal fluctuations alter immune and autonomic behavior across predictable cycles. Viral reactivation can raise baseline system load independently of external intervention [Kampa & Castanas, 2008; Cermakian et al., 2013].

Without incorporating these variables, intervention remains inconsistent across settings. With them, variability becomes measurable and actionable. The integration of phenotype, timing, delivery, and safety transforms variability from a confounder into a controllable parameter. This is the foundation required for predictive modeling. Once these relationships are defined, system behavior can be simulated rather than inferred, allowing intervention strategies to be optimized before they are applied.

9. Dynamic System Modeling and Predicted Outcomes 

To move from conceptual alignment to predictive capability, system behavior must be represented as a dynamic network of interacting variables. In immune-volatile systems, these variables are not independent. They are coupled through feedback relationships that determine whether the system amplifies or regulates incoming input.

System state can be expressed as:

S(t) = w₁N(t) + w₂I(t) + w₃A(t) + w₄C(t)

Each domain evolves over time according to its interaction with others:

dN/dt ∝ I + A + C dI/dt ∝ N + E + A dA/dt ∝ I + E + C dC/dt ∝ N + I + A

These relationships describe a system in which instability is self-reinforcing. Increased inflammation drives autonomic disruption. Autonomic disruption alters perfusion and signaling. Central sensitization amplifies perceived and physiological responses. Without intervention, the system tends toward persistence rather than resolution. These dynamics are consistent with network-based models of chronic disease and neuroimmune feedback loops [Barabási et al., 2011; Borsini et al., 2015; Morris et al., 2021].

Intervention as System Input

CRISPR intervention introduces an external input into this system:

S′(t) = S(t) + U(t) − R(t)

Where U(t) represents the magnitude of intervention and R(t) represents regulatory capacity.

The effect of intervention is therefore conditional. If U(t) is introduced when regulatory capacity is high and baseline load is low, the system absorbs and integrates the intervention. Editing occurs within a stable environment, increasing the likelihood of durable effect.

If the same U(t) is introduced when regulatory capacity is constrained, the system cannot process the input efficiently. This may lead to reduced editing efficiency, increased immune activation, or amplification of existing instability. In this context, what appears to be therapeutic failure is often a function of timing. This conditional response behavior is consistent with systems biology models in which intervention outcome depends on system position within a dynamic state space [Kitano, 2002].

Stabilization as a Control Mechanism

Stabilization alters both sides of this equation:

L(t) ↓ R(t) ↑

This dual effect is critical. Reducing load alone is insufficient if regulatory capacity remains impaired. Increasing regulatory capacity alone is insufficient if load remains high. STAIR operates by shifting both simultaneously, moving the system into a region where intervention can be integrated. This allows probability of success to be expressed as:

P(success) ∝ R(t) / L(t)

This relationship aligns with threshold-based models of physiologic resilience and adaptive capacity [Sterling, 2012].

Graded Input and Threshold Protection

Graded exposure modifies the magnitude of intervention input over time:

U₁ < U₂ < U₃ < U₄

Maintaining:

U(t) + L(t) < T

ensures that intervention does not exceed system tolerance at any stage. This is particularly important for CRISPR systems, where delivery involves not only molecular targeting but also interaction with immune recognition pathways and intracellular machinery [Pardi et al., 2018].

By introducing input progressively, the system adapts while maintaining stability. This improves both safety and interpretability, allowing true therapeutic effects to be distinguished from overload responses.

Predicted Outcome States

Within this model, outcomes are determined by the relationship between system load and regulatory capacity at the time of intervention.

High load / low regulation (L↑, R↓) Intervention amplifies instability. Adverse events increase, and efficacy appears reduced or inconsistent.

Moderate load / moderate regulation Partial integration occurs. Some benefit is observed, but variability remains high.

Low load / high regulation (L↓, R↑) Intervention integrates efficiently. Variability decreases, and therapeutic effects become more durable.

These outcome states are consistent with nonlinear system behavior observed in immune and neurobiological regulation [McEwen, 2007].

Interpretation

This framework demonstrates that variability in CRISPR outcomes is not an inherent limitation of the technology. It is a consequence of introducing intervention into systems operating at different levels of stability. By aligning intervention with system readiness, it becomes possible to reduce variability without changing the molecular target. This shifts the focus from improving targeting alone to optimizing the conditions under which targeting occurs. In doing so, CRISPR therapeutics move closer to achieving consistent, reproducible outcomes in complex biological systems.

Real-World Modeled Case: State-Dependent CRISPR Delivery in an Immune-Volatile System

To illustrate the practical implications of state-dependent intervention, consider a modeled patient with persistent viral activity consistent with EBV-driven immune dysregulation and a mixed phenotype characterized by inflammatory dominance, mast cell reactivity, and moderate autonomic instability. At baseline, system state is elevated across multiple domains. Inflammatory signaling remains persistently high due to viral activity, mast cell activation contributes to fluctuating reactivity, and environmental exposure further increases total system load. This can be represented as:

L₀ = I(7) + A(5) + M(6) + G(4) + E(6) + H(3) + V(7) ≈ High Load

At this level, L₀ approaches or exceeds the system’s tolerance threshold T. Under these conditions, direct introduction of CRISPR intervention results in a system response dominated by instability rather than integration. This can be expressed as:

U(t) + L₀ ≥ T

The predicted outcome in this state includes:

• increased likelihood of adverse reaction

• reduced effective uptake or engagement

• inconsistent or attenuated therapeutic effect

In standard trial design, this outcome would likely be classified as intolerance or lack of efficacy. Within a state-dependent framework, it is recognized as a mismatch between intervention and system readiness. Following STAIR-based stabilization, system load is reduced across domains. Inflammatory signaling decreases, mast cell activity stabilizes, environmental exposure is minimized, and autonomic variability is partially regulated. The system shifts toward:

L₁ = I(4) + A(3) + M(3) + G(3) + E(2) + H(3) + V(6) ≈ Moderate Load

At the same time, regulatory capacity increases, effectively raising T. This produces a new condition in which:

L₁ < T

Within this stabilized state, a viable intervention window emerges, consistent with:

W(t) > θ and dW/dt < 0

At this point, CRISPR intervention is introduced using a graded input model:

U₁ < U₂ < U₃ < U₄

Each stage is separated by sufficient time to observe delayed system response and ensure that:

U(t) + L₁ < T

This controlled introduction prevents threshold crossing and allows the system to integrate intervention progressively. The predicted outcome in this condition differs substantially from the unstabilized case:

• improved tolerability across stages

• more consistent cellular engagement

• reduced variability in response

• increased probability of durable therapeutic effect

Importantly, the molecular intervention itself has not changed. The difference in outcome is entirely attributable to alignment between intervention and system state. This contrast demonstrates a central principle of the framework. Variability in CRISPR outcomes is not solely a function of targeting or delivery technology. It is a function of when and how intervention is introduced relative to system conditions.

In unstabilized systems, therapeutic input competes with existing load and amplifies instability. In stabilized systems, the same input is processed within available regulatory capacity, allowing targeted effects to emerge. This distinction has direct implications for both clinical care and trial design. Without state-dependent alignment, effective therapies may be discarded due to inconsistent results. With alignment, those same therapies can demonstrate reproducible benefit.

State-Dependent Outcome Comparison 

To clarify the difference between unstabilized and stabilized intervention conditions, system behavior can be compared across load, regulatory capacity, and resulting outcomes.

Condition	System Load L(t)	Regulatory Capacity R(t)	Threshold Margin (T − L)	Intervention Behavior	Observed Outcome
Unstabilized	High (multi-domain elevation)	Constrained	Minimal or negative	Input competes with existing load	Amplified instability, adverse reactions, inconsistent engagement
Partially Stabilized	Moderate (reduced but variable)	Improving	Narrow positive margin	Input intermittently integrated	Mixed response, partial benefit, continued variability
Fully Stabilized (STAIR + Timing + Graded Input)	Lower, controlled	Elevated	Clear positive margin	Input processed within regulatory capacity	Consistent uptake, improved tolerability, durable response

Interpretation

This comparison highlights that intervention outcome is not determined solely by the therapeutic itself, but by the relationship between system load and regulatory capacity at the time of delivery [Kitano, 2002; Barabási et al., 2011]. In the unstabilized condition, high load compresses the threshold margin, leaving little room for additional input. Under these conditions, even precisely targeted CRISPR intervention behaves as a destabilizing force, as system-level stress and constrained adaptive capacity increase the likelihood of amplified response to additional biologic input [McEwen, 2007; Sterling, 2012]. The resulting adverse reactions and inconsistent engagement are frequently misinterpreted as therapeutic failure, even when they may instead reflect host-state mismatch rather than lack of molecular efficacy [Naldini, 2015; June et al., 2018]. In the partially stabilized condition, improvements in load and regulatory capacity allow for intermittent integration of intervention. 

However, because the threshold margin remains narrow, variability persists and outcomes remain inconsistent, which is consistent with nonlinear system behavior near threshold states [Borsini et al., 2015; Morris et al., 2021]. In the fully stabilized condition, reduction of system load combined with increased regulatory capacity creates a meaningful threshold margin. This allows graded intervention to remain within system tolerance at each stage, enabling consistent uptake and more durable therapeutic effects [Castells, 2017; Pardi et al., 2018]. This interpretation is also consistent with CYNAERA’s state-dependent frameworks, including STAIR and the Input Load Collapse Model, which treat response variability as a function of timing, threshold, and system readiness rather than as background noise alone.

Key Insight

The transition from failure to success does not require a change in the molecular intervention. It requires a shift in system conditions such that:

R(t) > L(t) and U(t) + L(t) < T

When these conditions are met, intervention is no longer competing with instability. It is integrated into a system capable of maintaining equilibrium.

10. Limitations and Translational Constraints

While the framework presented integrates system-state dynamics with CRISPR-based intervention, several limitations must be acknowledged. First, the model relies on the ability to approximate system variables such as inflammatory load, autonomic instability, and mast cell activity in real time. In current clinical settings, these variables are often inferred through proxy measures rather than directly quantified.

This introduces variability in how accurately system state can be assessed at the point of care [Benarroch, 2012; Raj et al., 2021]. Second, the equations presented describe relationships between domains rather than fully parameterized predictive models. While they reflect biologically grounded interactions, further work is required to formalize these relationships into quantitative simulations with validated coefficients, sensitivity analyses, and prospective testing across patient populations [Kitano, 2002; Barabási et al., 2011]. Third, delivery constraints remain a critical barrier. Achieving consistent CRISPR delivery to relevant tissues, including peripheral nerve structures and immune-active compartments, continues to present technical challenges. 

In addition, immune responses to delivery vectors may vary significantly across individuals, particularly in populations with baseline immune dysregulation [Lino et al., 2018; Pardi et al., 2018; Mitchell et al., 2021]. Fourth, the framework assumes the feasibility of implementing stabilization protocols such as STAIR prior to intervention. In practice, access to environmental control, medication support, and structured monitoring varies widely across healthcare settings. 

These conditions may limit the immediate scalability of the model [Marmot et al., 2008]. Finally, while the framework is designed to generalize across immune-volatile conditions, disease-specific factors may alter system dynamics in ways not fully captured here. Application to conditions beyond those modeled will require iterative validation and adaptation [Proal & VanElzakker, 2021]. These limitations do not negate the framework. They define the boundaries within which it must be tested, refined, and implemented. This framing is also consistent with CYNAERA’s staged modeling logic, in which system-state architecture is treated as a translational scaffold requiring iterative calibration rather than a finished universal simulation.

11. Safety, Adverse Events, and Ethical Considerations

The application of CRISPR-based therapeutics in immune-volatile populations requires a redefinition of safety that extends beyond traditional dose-dependent toxicity models. In these systems, adverse reactions are not determined solely by the therapeutic input itself, but by the interaction between that input and the system’s current load, regulatory capacity, and environmental context. This relationship can be expressed as:

Reaction Risk ∝ (U(t) + L(t)) / T

Where: U(t) = therapeutic input L(t) = cumulative system load T = tolerance threshold

Under this model, identical interventions may produce divergent outcomes depending on baseline system conditions. This is not variability without cause. It is threshold-dependent behavior [McEwen, 2007; Sterling, 2012].

11.1 Failure of Current Safety Recognition Models

Current clinical safety frameworks rely heavily on visible markers such as rash, swelling, or airway compromise. In mast cell-reactive and post-infectious populations, this assumption fails. Clinically significant reactions frequently present without cutaneous signs and instead manifest through blood pressure drops, tachycardia or autonomic instability, dizziness or near-syncope, gastrointestinal distress, and neurologic symptoms such as confusion or agitation [Afrin et al., 2016; Akin, 2017; Weinstock et al., 2021]. These reactions are routinely dismissed or misclassified in acute care settings, particularly in emergency departments. Patients are frequently labeled as anxious or non-compliant when objective physiologic changes are present but not externally visible. This creates two systemic failures: patient harm due to delayed or withheld intervention, and data distortion due to underreporting and misclassification of adverse events. In clinical trials, this leads to inaccurate safety profiles and contributes to the premature dismissal of viable therapies.

11.2 State-Dependent Safety as a Structured Process

To correct this, safety must be reframed as a structured, state-dependent process rather than a binary outcome. The core requirement is to ensure that:

U(t) + L(t) < T at each stage of exposure

This cannot be achieved through standard dosing protocols alone. It requires staged introduction and real-time assessment of system readiness [Castells, 2017; Torres et al., 2019].

11.3 Graded Exposure Protocol

A practical implementation of this model is reflected in ultra-sensitive trial protocols developed in MCAS and IACC populations [Castells, 2017; Weinstock et al., 2021]. This approach introduces therapeutic input progressively:

Input₁ < Input₂ < Input₃ < Input₄

Each stage is separated by sufficient time to observe delayed system response.

Operational progression Stage 0: Reactive window screening Intervention is deferred unless the system is in a stable state, defined by: no recent flare or anaphylaxis stable sleep and autonomic baseline low environmental stress load

Stage 1: External exposure (skin testing) Minimal surface-level input is applied to detect immediate reactivity without systemic engagement.

Stage 2: Re-exposure under mild perturbation The same input is applied after light skin disruption to assess sensitivity under altered conditions.

Stage 3: Mucosal micro-exposure (tongue-tip test) Low-level exposure without full systemic load allows early detection of intolerance.

Stage 4: Micro-dose systemic introduction Sub-threshold ingestion ensures that total system load remains below tolerance.

Stage 5: Controlled escalation Dose increases are introduced gradually, contingent on maintained stability.

This progression is not excessive caution. It is a method for mapping system response before threshold is exceeded.

11.4 Environmental and Physiologic Modifiers of Safety

System load is not determined solely by internal biology. It is continuously shaped by environmental and physiologic inputs. This can be formalized through environmental load modeling:

Air-Stress Load ∝ PM2.5 + Ozone + Humidity + Heat Index + Pressure Change

As demonstrated in environmental health and exposure literature, shifts in these variables can raise baseline system load within hours, narrowing tolerance thresholds and increasing reaction risk [Kampa & Castanas, 2008; Mendell et al., 2011; Dales et al., 2008]. Similarly, internal modifiers such as hormonal fluctuations, viral reactivation, sleep disruption, and recent exposure to triggers such as mold, chemicals, or allergens can alter both L(t) and T in real time [Cermakian et al., 2013; Proal & VanElzakker, 2021]. This means that safety cannot be assessed independently of context. The same intervention may be safe under one set of conditions and destabilizing under another. This logic also aligns with CYNAERA’s VitalGuard-type environmental load modeling, which treats external exposure as a real-time contributor to intervention readiness.

11.5 Implications for CRISPR Delivery

For CRISPR therapeutics, these dynamics are especially relevant. Unlike conventional drugs, CRISPR systems interact with intracellular regulatory machinery, may trigger immune recognition pathways, and rely on delivery vectors that themselves can provoke response [Doudna & Charpentier, 2014; Hsu et al., 2014; Pardi et al., 2018]. In unstable systems, high baseline load may reduce effective uptake, increase immune-mediated clearance, and amplify adverse reactions [Naldini, 2015]. Graded, state-dependent introduction mitigates these risks by ensuring that therapeutic input is introduced under conditions where regulatory capacity is sufficient to process it.

11.6 Ethical and Trial Design Implications

The failure to account for state-dependent safety introduces structural bias into clinical trials. Participants with the highest baseline instability are more likely to experience reactions, more likely to be excluded, and less likely to complete trials. This disproportionately removes the very populations most in need of treatment and skews trial data toward more stable individuals. Incorporating stabilization phases, readiness gating, graded exposure, and stabilization-based re-challenge reduces this bias and improves both equity and data integrity [Oh et al., 2015; Dzau & Balatbat, 2018].

11.7 Redefining Safety

Safety in immune-volatile systems is not defined solely by the absence of adverse events. It is defined by the ability to introduce intervention without exceeding system threshold, the ability to interpret reactions within system context, and the ability to maintain therapeutic engagement without destabilization. Absence of visible symptoms does not indicate absence of biological response. Presence of reaction does not necessarily indicate therapeutic failure. When safety is evaluated within a state-dependent framework, both patient outcomes and clinical data become more accurate. This is consistent with the broader argument advanced across CYNAERA’s state-dependent intervention models, in which safety and efficacy are co-determined by timing, threshold position, and system readiness rather than by dose alone.

12. Economic Impact and System-Level Value

The integration of state-dependent intervention frameworks into CRISPR-based therapeutics alters the economic structure of both drug development and chronic disease management. In current models, variability in immune-volatile populations is treated as an unavoidable cost of biological complexity. This assumption drives larger trials, higher dropout rates, increased adverse event burden, and the premature abandonment of therapies that may be effective under appropriate conditions [DiMasi et al., 2016; Wouters et al., 2020]. Within the framework presented in this paper, a significant portion of this variability is not irreducible. It is the result of misalignment between intervention and system state. Correcting this misalignment produces two distinct categories of economic value:

Direct Economic Savings

derived from reduced inefficiency within existing treatment models

Expansion Economic Value

 derived from increasing the size and characteristics of the treatable population

The cumulative impact can therefore be expressed as:

Cumulative Economic Value (CEV) = Direct Savings + Expansion Value

To illustrate the scale of this effect, the following modeled scenario uses a moderate-size therapeutic program and a conservative chronic illness population.

12.1 Direct Economic Savings

Direct savings arise from reducing inefficiencies introduced when therapies are applied without regard to system readiness.

Trial Efficiency Gains

Aligning intervention with system state reduces variability at the point of application, improving signal detection and reducing required sample sizes [Kitano, 2002; Barabási et al., 2011].

Formula 

Trial Savings = (N_base − N_aligned) × Cost_per_participant

Modeled Example 

Conventional trial size = 240 participants State-aligned trial size = 180 participants Cost per participant = $85,000

Calculation 

Trial Savings = (240 − 180) × 85,000 Trial Savings = 60 × 85,000 

Trial Savings = $5,100,000

Adverse Event Reduction

Adverse events in immune-volatile populations are frequently driven by threshold exceedance rather than intrinsic drug toxicity. Stabilization and graded introduction reduce these events [Castells, 2017; Sterling, 2012].

Formula 

Adverse Event Savings = (AE_base − AE_aligned) × Cost_per_AE

Modeled Example 

Baseline adverse events = 52 State-aligned adverse events = 28 Cost per event = $12,500

Calculation 

Adverse Event Savings = (52 − 28) × 12,500 Adverse Event Savings = 24 × 12,500 

Adverse Event Savings = $300,000

Dropout Waste Reduction

Dropout reflects unrecovered investment and reduced statistical power, often driven by instability-related intolerance rather than therapeutic failure [Walton et al., 2022].

Formula 

Dropout Waste Reduction = (D_base − D_aligned) × Cost_per_dropout

Modeled Example 

Baseline dropout = 72 participants State-aligned dropout = 18 participants Cost per dropout = $40,000

Calculation 

Dropout Waste Reduction = (72 − 18) × 40,000 Dropout Waste Reduction = 54 × 40,000 

Dropout Waste Reduction = $2,160,000

Chronic Burden Reduction

Improved stability and remission reduce long-term healthcare utilization and productivity loss in chronic disease populations [Cutler, 2020; Choutka et al., 2022].

Formula 

Chronic Burden Reduction = Additional Responders × Annual Cost Offset per Patient

Modeled Example 

Treated population = 10,000 patients Additional durable response due to state alignment = 12% Annual offset per patient = $28,000

Calculation 

Additional Responders = 10,000 × 0.12 = 1,200 Chronic Burden Reduction = 1,200 × 28,000 

Chronic Burden Reduction = $33,600,000 annually

Total Direct Economic Savings

Direct Savings = Trial Savings + Adverse Event Savings + Dropout Waste Reduction + Chronic Burden Reduction

Calculation Direct Savings = $5,100,000 + $300,000 + $2,160,000 + $33,600,000 

Direct Savings = $41,160,000 annually

12.2 Expansion Economic Value

The larger economic impact emerges from expanding who can be treated, when they can be treated, and how severe a patient can be while remaining eligible for therapy. State-dependent oral CRISPR frameworks reduce instability barriers that currently exclude large segments of the population from treatment.

This includes: patients unable to tolerate standard dosing patients with high baseline severity patients excluded due to age or fragility By enabling intervention in these populations, the framework increases both access and economic return, particularly in high-cost chronic illness groups [Wittenberg et al., 2024].

Access Expansion Value

Patients who are currently excluded due to instability or intolerance may become eligible under a stabilization-first framework.

Formula 

Access Expansion Value = Newly Eligible Patients × Annual Value per Patient

Modeled Example 

Baseline treated population = 10,000 Eligibility expansion = 20% Newly eligible patients = 2,000 Annual value per patient = $28,000

Calculation 

Access Expansion Value = 2,000 × 28,000 

Access Expansion Value = $56,000,000 annually

Severity Expansion Value

Patients with higher baseline severity often carry greater healthcare and disability costs. If these patients become treatable, the economic return rises disproportionately.

Formula 

Severity Expansion Value = Newly Treatable Severe Patients × Annual Severe-Cost Offset

Modeled Example 

Newly treatable severe patients = 1,200 Annual cost offset per patient = $55,000

Calculation 

Severity Expansion Value = 1,200 × 55,000 

Severity Expansion Value = $66,000,000 annually

Age Range Expansion Value

State-dependent delivery may enable treatment in populations currently excluded due to fragility or intolerance, including younger and older patients.

Formula 

Age Expansion Value = Additional Patients × Annual Cost Offset

Modeled Example 

Additional patients across age groups = 800 Annual offset per patient = $24,000

Calculation 

Age Expansion Value = 800 × 24,000 

Age Expansion Value = $19,200,000 annually

Total Expansion Value

Expansion Value = Access Expansion Value + Severity Expansion Value + Age Expansion Value

Calculation Expansion Value = $56,000,000 + $66,000,000 + $19,200,000 

Expansion Value = $141,200,000 annually

12.3 Total Cumulative Economic Value

CEV = Direct Savings + Expansion Value

Calculation CEV = $41,160,000 + $141,200,000 

CEV = $182,360,000 annually

12.4 Interpretation

The economic impact of state-dependent CRISPR intervention extends beyond reducing inefficiency within existing models. It fundamentally alters the boundaries of treatment.

In this modeled scenario, direct savings alone exceed $41 million annually, largely through improved trial efficiency, fewer adverse events, reduced dropout, and lower long-term chronic care burden. The larger effect comes from expansion value, which adds more than $141 million annually by increasing eligibility, reaching higher-severity patients, and extending intervention into groups that are currently excluded.

This reframes variability from a fixed cost into a controllable parameter. Instead of increasing trial size and resource allocation to compensate for noise, the framework reduces noise at the point of intervention. The result is not only improved clinical outcomes, but a more efficient and scalable model for therapeutic development and deployment. For this reason, state-dependent CRISPR delivery should not be evaluated only as a scientific refinement. It should be understood as an economic architecture that converts biologic instability from a source of waste into a source of measurable strategic value.

13. Conclusion

The development of CRISPR-based therapeutics represents a significant advance in the ability to target disease-driving pathways [Doudna & Charpentier, 2014]. However, precision at the molecular level does not guarantee consistent clinical outcomes in systems characterized by immune volatility, multi-domain interaction, and dynamic instability. This paper argues that the primary barrier to translation is not solely technical. It is architectural. Current models apply intervention without accounting for system state, resulting in variability that is misinterpreted as biological unpredictability or therapeutic limitation. By introducing a state-dependent framework, this variability can be reframed as a function of measurable relationships between system load, regulatory capacity, and intervention timing [Kitano, 2002; Barabási et al., 2011].

This framework establishes that system behavior can be represented as a dynamic interaction of coupled domains, intervention outcomes depend on alignment between input and system state, stabilization and timing modify both safety and efficacy, and graded introduction preserves system integrity while improving interpretability. The modeled case demonstrates that identical CRISPR interventions can produce divergent outcomes depending on whether they are introduced into stabilized or unstable systems. This distinction highlights a critical insight: variability in outcome is not random. It is predictable.

This has direct implications for clinical practice and research. Incorporating system-state awareness into trial design, delivery strategy, and safety evaluation can improve reproducibility without requiring changes to molecular targeting alone. It shifts the focus from refining intervention in isolation to optimizing the conditions under which intervention occurs. Importantly, this shift also carries significant economic implications. As demonstrated in Section 12, a substantial portion of cost in current therapeutic development arises from variability that is treated as unavoidable but is in fact driven by misalignment between intervention and system state.

By reducing this misalignment, state-dependent frameworks decrease trial size requirements, lower adverse event burden, reduce dropout-related losses, and prevent premature abandonment of viable therapies [DiMasi et al., 2016; Wouters et al., 2020]. At the same time, they expand the treatable population by enabling intervention in patients who are currently excluded due to instability, intolerance, or severity. The combined effect is structural, shifting both the cost curve and the accessible market for CRISPR-based therapies.

More broadly, this framework suggests that the future of gene-based therapeutics in complex disease will not be defined solely by advances in editing technology. It will be defined by the ability to integrate those technologies within dynamic biological systems. In that context, remission is no longer an abstract goal. It becomes an engineering problem grounded in alignment, timing, and system capacity. Within this paradigm, CRISPR Remission™ represents an early applied implementation of state-dependent gene editing architecture, integrating flare-aware timing, environmental modulation, and biologic readiness into therapeutic design. Its acceptance at the CRISPR Medicine Conference (CRISPRMED26) reflects growing recognition that delivery context, not just molecular targeting, determines clinical outcome. CYNAERA’s state-dependent modeling systems, positions remission not as a single mechanistic event, but as the result of coordinated intervention within a biologically readable system. When variability is understood as a function of system state rather than noise, both clinical outcomes and economic performance become optimizable within the same framework.

Frequently Asked Questions

What is state-dependent CRISPR delivery?

State-dependent CRISPR delivery refers to deploying gene editing interventions based on the patient’s real-time biological state rather than using a fixed, one-time treatment model. In CYNAERA’s framework, delivery is aligned with immune activity, environmental burden, and flare dynamics to improve both safety and durability of response.

Why is CRISPR delivery difficult in real-world conditions?

CRISPR delivery remains one of the primary challenges in gene editing. Current approaches rely on viral, chemical, or physical delivery systems, each with tradeoffs in targeting precision, immune response, and distribution within the body. In dynamic diseases, these challenges are amplified because the biological environment is constantly shifting, which can alter how therapies are absorbed, distributed, and tolerated.

What makes oral CRISPR delivery different from traditional delivery methods?

Oral CRISPR delivery introduces additional complexity because therapies must pass through the gastrointestinal system, where factors such as permeability, immune activation, and microbiome interactions influence effectiveness. Unlike localized or injected delivery, oral approaches are highly sensitive to environmental and physiologic conditions, making them inherently state-dependent.

How do environmental factors affect CRISPR delivery?

Environmental exposures such as air quality, mold, toxins, and chemical irritants can alter immune signaling, inflammation, and tissue stability. In CYNAERA’s framework, these factors are treated as part of the delivery equation, not external variables. Environmental burden can influence whether CRISPR therapies are tolerated, effective, or destabilizing at the time of intervention.

What is flare-aware CRISPR intervention?

Flare-aware intervention means timing CRISPR delivery to periods of lower biological volatility, often referred to as stabilization windows. In immune-volatile conditions, delivering therapy during a flare may increase risk and reduce effectiveness, while delivery during stabilization can improve outcomes and reduce rebound effects.

Can CRISPR be used to target viral persistence such as EBV?

CRISPR-based approaches are being explored for targeting persistent viral reservoirs, including Epstein-Barr virus (EBV), by editing or suppressing viral DNA within host cells. However, success depends not only on targeting accuracy, but also on the surrounding biological environment, including immune state and inflammatory signaling.

Why do current CRISPR delivery models struggle with chronic illness?

Most CRISPR delivery systems are designed for biologically stable conditions, where cell behavior and immune response are relatively predictable. Chronic and post-infectious conditions are different. They involve fluctuating immune states, environmental sensitivity, and multi-system interaction, which can disrupt delivery efficiency and treatment durability.

How does CYNAERA improve CRISPR delivery outcomes?

CYNAERA’s approach integrates delivery strategy with system-state modeling, including:

• immune activity and inflammatory load

• environmental exposure

• flare timing and recovery cycles

This allows CRISPR delivery to be treated as a dynamic process rather than a static event, improving both safety and long-term effectiveness.

Is CRISPR delivery only a technical problem?

No. While delivery technologies such as viral vectors, nanoparticles, and exosomes are advancing, the main limitation is not just how CRISPR is delivered, but when and under what conditions it is delivered. CYNAERA reframes delivery as a system-level problem involving timing, environment, and patient-specific biology, not just engineering.

CYNAERA Framework Papers

This paper draws on a defined subset of CYNAERA Institute white papers that establish the methodological and analytical foundations of CYNAERA’s frameworks. These publications provide deeper context on prevalence reconstruction, remission, combination therapies and biomarker approaches. Our Long COVID,  ME/CFS  and CRISPR Remission Libraries are also in depth resources.

• Primary Chronic Trigger 

• Mold Exposure as A Flare Catalyst in ME/CFS

• VitalGuard™: Environmental Flare Risk Engine

• Corrected National Prevalence Estimates for (IACCs)

• Undercounted from Kabul to Kansas: The Hidden Men of IACC

• The Science of Remission of IACC's

• A Nobel-Scale Advance: AI-Powered CRISPR Platform to End IACC's

• State-Dependent CRISPR Delivery

Author’s Note:

All insights, frameworks, and recommendations in this written material reflect the author's independent analysis and synthesis. References to researchers, clinicians, and advocacy organizations acknowledge their contributions to the field but do not imply endorsement of the specific frameworks, conclusions, or policy models proposed herein. This information is not medical guidance.

Patent-Pending Systems

​Bioadaptive Systems Therapeutics™ (BST) and all affiliated CYNAERA frameworks, including CRISPR Remission™, VitalGuard™, CRATE™, SymCas™, and RAVYNS™, are protected under U.S. Provisional Patent Application No. 63/909,951.

Licensing and Integration

CYNAERA partners with universities, research teams, federal agencies, health systems, technology companies, and philanthropic organizations. Partners can license individual modules, full suites, or enterprise architecture. Integration pathways include research co-development, diagnostic modernization projects, climate-linked health forecasting, and trial stabilization for complex cohorts. You can get basic licensing here at CYNAERA Market.

Support structures are available for partners who want hands-on implementation, long-term maintenance, or limited-scope pilot programs.

About the Author 

Cynthia Adinig is a researcher, health policy advisor, author, and patient advocate. She is the founder of CYNAERA and creator of the patent-pending Bioadaptive Systems Therapeutics (BST)™ platform. She serves as a PCORI Merit Reviewer, and collaborator with Selin Lab for T cell research at the University of Massachusetts.

Cynthia has co-authored research with Harlan Krumholz, MD, Dr. Akiko Iwasaki, and Dr. David Putrino, though Yale’s LISTEN Study, advised Amy Proal, PhD’s research group at Mount Sinai through its patient advisory board, and worked with Dr. Peter Rowe of Johns Hopkins on national education and outreach focused on post-viral and autonomic illness. She has also authored a Milken Institute essay on AI and healthcare, testified before Congress, and worked with congressional offices on multiple legislative initiatives. Cynthia has led national advocacy teams on Capitol Hill and continues to advise on chronic-illness policy and data-modernization efforts.

Through CYNAERA, she develops modular AI platforms, including the CRISPR Remission™, IACC Progression Continuum™, Primary Chronic Trigger (PCT)™, RAVYNS™, and US-CCUC™, that are made to help governments, universities, and clinical teams model infection-associated conditions and improve precision in research and trial design. US-CCUC™ prevalence correction estimates have been used by patient advocates in congressional discussions related to IACC research funding and policy priorities. Cynthia has been featured in TIME, Bloomberg, USA Today, and other major outlets, for community engagement, policy and reflecting her ongoing commitment to advancing innovation and resilience from her home in Northern Virginia.

Cynthia’s work with complex chronic conditions is deeply informed by her lived experience surviving the first wave of the pandemic, which strengthened her dedication to reforming how chronic conditions are understood, studied, and treated. She is also an advocate for domestic-violence prevention and patient safety, bringing a trauma-informed perspective to her research and policy initiatives.

References

1. Akin, C. (2017). Mast cell activation syndromes. Journal of Allergy and Clinical Immunology, 140(2), 349–355. https://doi.org/10.1016/j.jaci.2017.06.007

2. Barabási, A.-L., Gulbahce, N., & Loscalzo, J. (2011). Network medicine: A network-based approach to human disease. Nature Reviews Genetics, 12(1), 56–68. https://doi.org/10.1038/nrg2918

3. Benarroch, E. E. (2020). Physiology and pathophysiology of the autonomic nervous system. Continuum (Minneap Minn), 26(1), 12–24. https://doi.org/10.1212/CON.0000000000000815

4. Borsini, A., Zunszain, P. A., Thuret, S., & Pariante, C. M. (2015). The role of inflammatory cytokines as key modulators of neurogenesis. Trends in Neurosciences, 38(3), 145–157. https://doi.org/10.1016/j.tins.2014.12.006

5. Castells, M. (2009). Rapid desensitization for hypersensitivity reactions to medications. Immunology and Allergy Clinics of North America, 29(3), 585–606. https://doi.org/10.1016/j.iac.2009.04.012

6. Cermakian, N., Lange, T., Golombek, D., Sarkar, D. K., Nakao, A., Shibata, S., & Mazzoccoli, G. (2013). Crosstalk between the circadian clock circuitry and the immune system. Chronobiology International, 30(7), 870–888. https://doi.org/10.3109/07420528.2013.782315

7. Choutka, J., Jansari, V., Hornig, M., & Iwasaki, A. (2022). Unexplained post-acute infection syndromes. Nature Medicine, 28, 911–923.

8. Cutler, D. M. (2020). The costs of chronic disease in the United States. JAMA.

9. Dales, R., Liu, L., Szyszkowicz, M., et al. (2008). Particulate air pollution and vascular reactivity. Environmental Health Perspectives.

10. DiMasi, J. A., Grabowski, H. G., & Hansen, R. W. (2016). Innovation in the pharmaceutical industry: New estimates of R&D costs. Journal of Health Economics, 47, 20–33.

11. Doudna, J. A., & Charpentier, E. (2014). The new frontier of genome engineering with CRISPR-Cas9. Science, 346(6213), 1258096. https://doi.org/10.1126/science.1258096

12. Hsu, P. D., Lander, E. S., & Zhang, F. (2014). Development and applications of CRISPR-Cas9 for genome engineering. Cell, 157(6), 1262–1278.

13. June, C. H., O’Connor, R. S., Kawalekar, O. U., Ghassemi, S., & Milone, M. C. (2018). CAR T cell immunotherapy for human cancer. Science, 359(6382), 1361–1365.

14. Kampa, M., & Castanas, E. (2008). Human health effects of air pollution. Environmental Pollution, 151(2), 362–367. https://doi.org/10.1016/j.envpol.2007.06.012

15. Kieser, A., & Sterz, K. R. (2015). The latent membrane protein 1 (LMP1). Current Topics in Microbiology and Immunology, 391, 119–149. https://doi.org/10.1007/978-3-319-22834-1_4

16. Kitano, H. (2002). Systems biology: A brief overview. Science, 295(5560), 1662–1664. https://doi.org/10.1126/science.1069492

17. Lino, C. A., Harper, J. C., Carney, J. P., & Timlin, J. A. (2018). Delivering CRISPR: A review of the challenges and approaches. Drug Delivery, 25(1), 1234–1257. https://doi.org/10.1080/10717544.2018.1474964

18. Ma, Y., Walsh, M. J., Bernhardt, K., et al. (2017). CRISPR/Cas9 screens reveal Epstein-Barr virus-transformed B cell host dependency factors. Cell Host & Microbe, 21(5), 580–591.e7. https://doi.org/10.1016/j.chom.2017.04.010

19. Marmot, M., Friel, S., Bell, R., Houweling, T. A. J., & Taylor, S. (2008). Closing the gap in a generation: Health equity through action on the social determinants of health. The Lancet, 372(9650), 1661–1669. https://doi.org/10.1016/S0140-6736(08)61690-6

20. McEwen, B. S. (2007). Physiology and neurobiology of stress and adaptation: Central role of the brain. Physiological Reviews, 87(3), 873–904. https://doi.org/10.1152/physrev.00041.2006

21. Mendell, M. J., Mirer, A. G., Cheung, K., Tong, M., & Douwes, J. (2011). Respiratory and allergic health effects of dampness, mold, and dampness-related agents: A review of the epidemiologic evidence. Environmental Health Perspectives, 119(6), 748–756. https://doi.org/10.1289/ehp.1002410

22. Mitchell, M. J., Billingsley, M. M., Haley, R. M., Wechsler, M. E., Peppas, N. A., & Langer, R. (2021). Engineering precision nanoparticles for drug delivery. Nature Reviews Drug Discovery, 20(2), 101–124. https://doi.org/10.1038/s41573-020-0090-8

23. Morris, G., Puri, B. K., Olive, L., et al. (2021). Psychiatric symptoms and neuroimmune mechanisms in COVID-19 and long COVID. Molecular Neurobiology.

24. Naldini, L. (2015). Gene therapy returns to centre stage. Nature, 526(7573), 351–360. https://doi.org/10.1038/nature15818

25. Pardi, N., Hogan, M. J., Porter, F. W., & Weissman, D. (2018). mRNA vaccines, a new era in vaccinology. Nature Reviews Drug Discovery, 17(4), 261–279. https://doi.org/10.1038/nrd.2017.243

26. Proal, A. D., & VanElzakker, M. B. (2021). Long COVID or post-acute sequelae of COVID-19 (PASC): An overview of biological factors that may contribute to persistent symptoms. Frontiers in Microbiology, 12, 698169. https://doi.org/10.3389/fmicb.2021.698169

27. Sterling, P. (2012). Allostasis: A model of predictive regulation. Physiology & Behavior, 106(1), 5–15. https://doi.org/10.1016/j.physbeh.2011.06.004

28. Torres, M. J., Madrigal-Burgaleta, R., et al. (2019). Hypersensitivity reactions to biologics. Journal of Investigational Allergology and Clinical Immunology.

29. Walton, M. K., Powers, J. H., Hobart, J., et al. (2022). Clinical outcome assessments and patient dropout considerations in trial design. Value in Health.

30. Weinstock, L. B., Brook, J. B., Walters, A. S., Goris, A., Afrin, L. B., Molderings, G. J., et al. (2021). Mast cell activation symptoms are prevalent in Long-COVID. International Journal of Infectious Diseases, 112, 217–226. https://doi.org/10.1016/j.ijid.2021.09.043

31. Wittenberg, E., Choi, Y. K., et al. (2024). Economic burden and quality-of-life losses in Long COVID and related chronic illness populations. Medical Care.

32. Wouters, O. J., McKee, M., & Luyten, J. (2020). Estimated research and development investment needed to bring a new medicine to market, 2009–2018. JAMA, 323(9), 844–853. https://doi.org/10.1001/jama.2020.1166

33. Young, L. S., Yap, L. F., & Murray, P. G. (2016). Epstein-Barr virus: More than 50 years old and still providing surprises. Nature Reviews Cancer, 16(12), 789–802.