ApoE Mimetic Peptides as Therapy for Traumatic Brain Injury

Abstract

The lack of targeted therapies for traumatic brain injury (TBI) remains a compelling clinical unmet need. Although knowledge of the pathophysiologic cascades involved in TBI has expanded rapidly, the development of novel pharmacological therapies has remained largely stagnant. Difficulties in creating animal models that recapitulate the different facets of clinical TBI pathology and flaws in the design of clinical trials have contributed to the ongoing failures in neuroprotective drug development. Furthermore, multiple pathophysiological mechanisms initiated early after TBI that progress in the subacute and chronic setting may limit the potential of traditional approaches that target a specific cellular pathway for acute therapeutic intervention. We describe a reverse translational approach that focuses on translating endogenous mechanisms known to influence outcomes after TBI to develop druggable targets. In particular, numerous clinical observations have demonstrated an association between apolipoprotein E (apoE) polymorphism and functional recovery after brain injury. ApoE has been shown to mitigate the response to acute brain injury by exerting immunomodulatory properties that reduce secondary tissue injury as well as protecting neurons from excitotoxicity. CN-105 represents an apoE mimetic peptide that can effectively penetrate the CNS compartment and retains the neuroprotective properties of the intact protein.

Introduction

Traumatic brain injury (TBI) is an acquired condition that results from an external mechanical force applied to the skull and underlying intracranial contents resulting in temporary or permanent damage to affected brain tissue or function [1]. Globally, an estimated 64–74 million people sustain a TBI every year, the majority of which are classified as mild TBI, also commonly referred to as concussion [2, 3]. TBIs are associated with numerous mechanisms including sports-related injuries, falls, motor vehicle accidents, and assaults. Additionally, factors such as location, socioeconomic status, and biological sex can impact TBI incidence and outcome [4]. Between 2001 and 2012 in the USA, an estimated 3.42 million sports- and recreation-related TBIs were evaluated in US emergency departments [5]. According to the US Centers for Disease Control and Prevention, there were approximately 223,135 TBI-related hospitalizations in 2019, and 64,362 TBI-related deaths in 2020 [6]. For certain populations, additional mechanisms of brain injury exist, such as military personnel who may be exposed to blast and direct energy sources [7–9]. Between 2001 and 2021, more than 1.6 million military personnel deployed to Iraq or Afghanistan in support of the Global War on Terror (GWOT), with an estimated 5 to 35% of them having suffered a concussion [7]. As of August 2022, the Traumatic Brain Injury Center of Excellence (TBICoE) reports 463,392 TBIs across the Department of Defense since 2000, 82% of which are mild in severity [10]. Despite improvements in the recognition and documentation of concussions, it is widely believed that the incidence is drastically underestimated, with as few as 1 out of every 9 concussions being captured by current data collection methods [1, 5]. Practically, TBI can result in significant long-term cognitive and physical sequelae that have a significant impact on the individual’s personal and professional life, as well as the society and economy. Annual costs, both direct and indirect, have previously been calculated to total more than $60 billion annually in the USA [11]. As a result, there is significant interest in improving diagnostic accuracy and therapeutic options for TBI.

TBI Therapeutic Landscape: Unmet Needs

A combination of unknown incidence, poorly understood pathophysiology, historically unrecognized or misattributed signs and symptoms, and the paucity of options to aid in the diagnosis and treatment of TBI have led some to refer to it as a “silent epidemic” [12]. Between the increased attention to the long-term effects of head injury in sports and the prevalence of blast TBI in the GWOT over the last 20 years, there has been steadily increasing awareness of the effects and cost of TBI among the civilian and military populations [7, 13]. Indeed, significant progress has been made in many domains including public health initiatives that have led to seatbelt legislation, improved workplace and safety regulations, improved helmet design for sports and the military, effective screening tools and protocolized approaches to assessment and care, and return to activity recommendations for sports. The establishment of specialized neurocritical care units and the focus on specialized TBI rehabilitation have also contributed to improved outcomes, particularly in patients with moderate and severe TBI [7, 14]. Despite these advances, an important area where progress remains elusive is the development of effective neuroprotective agents.

Although decades of research have yielded pharmacological interventions with promising preclinical results, none have demonstrated benefit in improving long-term outcomes once they reached Phase 3 clinical trials. This 100% failure rate suggests a problem with study design at multiple levels. More specifically, criticisms of TBI trials can be divided into problems with preclinical testing or clinical trial design and testing [15]. Problematic issues in the preclinical setting include inadequate attention to the temporal profile of specific pathobiological responses that contribute to secondary injury [16]. For example, both the mechanical stretch and secondary ischemia associated with the primary concussive force lead to overactivation of glutamate receptors, resulting in neuronal excitotoxic injury [17]. The temporal window for therapeutic intervention to mitigate this injury cascade is brief compared to the neuroinflammatory response, which is associated with diffuse glial activation, leading to secondary tissue injury and cerebral edema, and may progress over a period of days [18–20].

In TBI studies, it is important to recognize the biological differences between human and rodent brain structure and function, and that laboratory methods often fail to replicate the heterogenous injury patterns that occur with real-world TBI [21–26]. The known anatomic differences, such as the absence of gyri and a small amount of white matter in the rodent brain, as well as geometric variations in the neuraxis of animal brains, limit the ability to produce injury patterns comparable to those in humans. Furthermore, many animal models focus on short-term functional outcomes and cannot reproduce emotional or language deficits that can occur in human TBI [14, 27]. Additional limitations associated with preclinical testing often include inadequate sample size [28] and inadequate blinding, randomization, reporting bias, attrition, and selective analysis [29–31]. Translational issues such as inadequate understanding of the pharmacokinetics, dosages, and differences between human and experimental animals may also limit the potential translatability of promising preclinical findings. Moreover, preclinical studies may not adequately address efficacy across, sex, age, and medical co-morbidities, which are needed to better assess the populations of patients who are most likely to benefit from therapies [14, 15, 32].

Methods for producing direct head trauma in animals generally produce homogenous injury patterns that are well-controlled by design [27]. The controlled cortical impact (CCI) method, for example, is one of the most widely utilized approaches because it can create consistent injuries associated with durable functional deficits [33, 34]. CCI is used post-craniotomy to produce mild to severe injuries via direct cortical deformation using a 3 mm or 5 mm impactor with a velocity of 3–6 m/s resulting in a displacement of the cortex of 0.5–1.5 mm [35]. Spatial learning and memory deficits, recognition, limbic and frontal dysfunction, and a variety of motor impairments can be reliably produced with CCI in rodents. Histologically, CCI can reproduce primary and secondary neuronal death, hippocampal and subventricular zone cell proliferation, blood–brain barrier disruption, and oxidative stress and neuroinflammation [33, 34]. While easy to use, reproducible, and capable of producing a range of injury severity, CCI can produce variable and sometimes unintended injuries. Recent data has shown that lateral movements of the piston delivering the cortical impact can increase the number of secondary impacts. While the animal skull is fixed in a stereotaxic frame to limit the potential for secondary impacts, this immobility risks increased tissue damage and injury severity [34]. Furthermore, CCI can lead to dural lacerations that create an open head injury, the type of impactor tip can affect the rate of neocortical degeneration, and the mild to moderate forces produced by this method may not induce brainstem injury necessary for severe TBI models [33, 35].

Importantly, although models that employ a targeted impact directly to brain tissue may create a standardized and homogenous injury, they are limited in their ability to recapitulate the biomechanical forces associated with clinical closed head injury [35], which are associated with complex pathology including parenchymal, subarachnoid, subdural and epidural hemorrhage, cortical contusion, and sheer injury. Although all lissencephalic rodent models of TBI imperfectly recapitulate the clinically relevant biomechanics seen in human closed head injury, we favor pneumatic impact models in which a controlled force and displacement are delivered stereotactically to the intact skull. These models allow for diffuse glial activation and injury to selectively vulnerable remote hippocampal and cortical populations and produce durable neurocognitive deficits that are modifiable by therapeutic intervention [29, 36, 37].

Limitations in clinical trial design for patients with closed head injuries have also played an important role in the failure of neuroprotective therapy development. These include the use of single-center randomized clinical trials (RCTs) [38], follow-up bias, inadequate or unclear random sequence generation or blinding with subjective endpoints [39], poor translation of doses from preclinical studies to clinical trials [40], poor patient selection [41], misclassification of outcomes reducing effect size (i.e., Glasgow Outcome Scale score) [42], and inadvertent bias due to external pressures [31]. Improvements in study design such as better patient selection criteria (such as stratification by serum biomarkers as well as age, sex, radiographic findings, and type of injury), adaptive designs that eliminate locked-in assignments in favor of changes in group assignments as the trial progresses, the inclusion of multiple quantitative outcome measures, optimization of dose and route administration, and a stronger understanding of drug mechanisms and metabolic characteristics may ultimately lead to a greater track record of success [32].

Reverse Translation: from Genetic Association to Therapy

Traditional drug development programs often utilize a reductionistic approach, whereby a specific cellular pathway believed to influence disease pathology is targeted with an optimized pharmacologic intervention. However, despite our increased understanding of the molecular and cellular mechanisms associated with secondary neuronal injury in the setting of acute CNS injury, no clinically effective neuroprotectant has been identified or developed. It is likely that this failure of translation is due, in part, to the pleiotropic effects of pathways involved in inflammation, oxidative stress, and excitotoxicity, as well as compensatory cellular mechanisms. An alternative to this traditional approach of “bench to bedside” is reverse translation. This “bedside to bench” approach focuses on the translation of an endogenous protein known to influence outcomes into a druggable target (Fig. 1). For example, functional outcome after traumatic brain injury is widely variable, and this has prompted the study of genetic influences on recovery [43]. Although several genetic polymorphisms have been demonstrated to influence outcomes, one of the leading candidates for this process of reverse translation is the apoE polymorphism, which encodes the human apolipoprotein E2, E3, and E4 protein isoforms. Although these common human isoforms only differ by single cysteine to arginine substitutions (apoE3: Cys 112 Arg 158; apoE4: Arg 112 Arg 158; apoE2: Cys 112 Cys 158), the cysteine to arginine substitution at position 112 is believed to cause domain interactions which result in structural alterations of apoE4 that influence its metabolism and functional activity [44, 45].

Fig. 1.

Translation vs. reverse translation. The conventional bench to bedside approach for development of therapeutic strategies, known a translation, involves basic science research in the lab, leading to drug development, and culminating in human clinical trials. Alternatively, a bedside to bench approach, known as reverse translation, takes clinical observations and utilizes them to inform the development of new therapies in the laboratory. Considering the slow progress in identifying an effective therapeutic for TBI using traditional methods thus far, reverse translation may augment traditional approaches by helping to identify therapeutic targets and influence trial design

Although initially identified in the context of cholesterol metabolism, in the 1990s apoE polymorphism was found to have a robust effect on the development of late-onset sporadic and familial Alzheimer’s disease [46]. In particular, the presence of one apoE4 allele resulted in an approximately fourfold increase in the development of AD, whereas homozygosity for apoE4 was associated with an approximately tenfold increase. Based on the identification of apoE4 as a genetic risk factor for AD, initial studies focused on disease-specific pathologies, such as isoform-specific effects on Aβ metabolism and deposition [47] and interactions with the microtubule-associated protein tau [48]. Interestingly, several studies have also demonstrated that the presence of apoE4 adversely affects outcomes following acute brain injury [49, 50]. Similarly, clinical and preclinical evidence suggests that apoE4 may be associated with impaired cognitive outcomes after mild TBI [51] and the development of cognitive dysfunction and chronic traumatic encephalopathy in older patients after repetitive head injuries [52–55], although the latter association remains somewhat controversial [56, 57].

As glial activation and neuroinflammatory responses play an important role in both acute brain injury responses and the development of neurodegenerative disease, one unifying hypothesis that would accommodate a role for apoE polymorphism in modifying outcomes across this spectrum of acute and chronic neurological disease is an isoform-specific effect on neuroinflammatory pathways. In fact, initial observations from more than 25 years ago have demonstrated a role for apoE in modifying glial activation and neuroinflammatory responses [58, 59]. Moreover, several preclinical and clinical studies have demonstrated isoform-specific effects on inflammation, with the apoE4 protein isoform associated with increased neuroinflammatory responses relative to apoE3 [60–64]. These isoform-specific effects on neuroinflammatory pathways may also extend to the blood–brain barrier permeability, which plays a particularly important role in the development of cerebral edema following traumatic brain injury [65]. A more complete understanding of the mechanism(s) of the isoform-specific effects by which apoE modifies neuroinflammatory mechanisms has important therapeutic implications and remains incompletely defined. ApoE binds a family of low-density lipoprotein (LDL) receptors, and observations suggest that interactions with the low-density lipoprotein receptor-related protein (LRP1) may initiate a signaling cascade [66–70] that are critical in mediating its immunomodulatory properties via mitogen-activated protein kinase (MAP kinase) pathways [71–73], and stabilization of the blood–brain barrier through the cyclophilin A-matrix metallopeptidase-9 pathway [74].

Despite the numerous clinical observations suggesting that the apoE isoform modifies functional outcomes after acute brain injury, clinical observational studies are often limited by the heterogeneity of injury and small population sizes. In this regard, the use of standardized injury paradigms in targeted replacement mice expressing the human apoE3 and apoE4 isoforms allows an important tool to validate the effect of apoE polymorphisms in the setting of acute brain injury, as well as to allow for the study of cellular mechanisms by which these effects are mediated.

Early preclinical studies have consistently demonstrated that apoE plays an important function in modulating systemic and CNS inflammatory responses [75–80]. ApoE4 mice have enhanced inflammatory responses as compared to their apoE3 counterparts in a variety of experimental paradigms. For example, after peripheral injection of bacterial endotoxin (lipopolysaccharide), there was enhanced evidence of peripheral and CNS pro-inflammatory responses, suggesting a dynamic interplay between peripheral and CNS inflammatory mediators [80]. Moreover, the presence of the apoE4 allele was associated with enhanced neuroinflammation and worsened outcomes in preclinical models of traumatic brain injury [81] and standardized models designed to mimic many of the pathological features of TBI, including subarachnoid hemorrhage [82] intraparenchymal hemorrhage [83] and cerebral ischemia [84, 85]. In total, these preclinical and clinical observations support the contention that apoE modifies acute brain injury responses in an isoform-specific fashion, in part via modulation of maladaptive neuroinflammatory responses [60].

Preclinical Development of CN-105, a Candidate apoE Mimetic Peptide

Based on the evidence that apoE3 plays an adaptive role following acute brain injury, a plausible therapeutic strategy would be to augment the beneficial effects of the endogenous protein. However, due to its size, peripherally administered apoE does not readily cross the blood–brain barrier, and in fact, the apoE produced primarily by astrocytes within the CNS compartment represents a relatively discrete pool from that found in the periphery, where synthesis occurs primarily in the liver [86]. However, dimerized peptides derived from the apoE receptor binding region had been demonstrated to have effects on lymphocyte proliferation [87–89]. Thus, to harness the therapeutic potential of apoE, a series of smaller peptides were created from the residues 130–149 of the apoE receptor binding region to explore the hypothesis that a smaller peptide capable of penetrating the CNS compartment could retain the receptor interactions necessary to mediate the bioactivity of the intact holoprotein [62, 90]. Although there was a size threshold by which the peptides lost the helicity necessary for bioactivity [62], the therapeutic peptide COG1410 represented a 12-amino acid peptide in which the helical structure was stabilized by two aminoisobutyric (Aib) residues that retained the anti-inflammatory effects of the apoE holoprotein [91].

As noted above, TBI is associated with heterogenous pathology, and to facilitate the translation of new therapeutic interventions, it is helpful to dissect the underlying pathobiological responses. The temporal profile of secondary tissue injury due to neuroinflammation may extend over several days, making it an attractive target for pharmacological intervention. Moreover, secondary tissue injury associated with neuroinflammation is common to many of the mechanisms of injury associated with TBI pathology, including ischemia, contusion, subarachnoid, and parenchymal hemorrhage [92]. The intracellular cascade associated with glutamate excitotoxicity is another mechanism of neuronal cell death following trauma and ischemia. In this regard, it is notable that in addition to its effects on microglial activation and neuroinflammation, apoE mimetic peptides also reduced excitotoxic neuronal injury in the same manner as the intact apoE [93], a property which was hypothesized to also be mediated by LRP-1 interactions [94, 95]. Thus, apoE mimetic peptides have pleiotropic effects on glutamate excitotoxicity and inflammation, and intravenous administration of the apoE mimetic peptide reduced histological injury and had a durable effect on improving functional outcomes in murine models of closed head injury [96, 97], as well as models of cerebral ischemia [98], and parenchymal and subarachnoid hemorrhage [82, 99, 100]. These models recapitulate different aspects of clinical TBI pathology [92].

Although well tolerated and associated with functional improvement, a limitation of COG1410 was its relatively large size, low potency, and the cost associated with the incorporation of non-naturally occurring Aib residues. To enhance druggability, a newer generation of apoE mimetic peptides was developed by modeling the polar receptor binding face of the amphipathic helical receptor binding region of apoE involved in direct receptor interactions. This library of peptides was tested in assays of glial activation [91] and excitatory [94] neuroprotection to generate the candidate apoE mimetic pentapeptide CN-105 (Ac-VSRRR-NH₂) (Fig. 2). To date, CN-105 has demonstrated histological and functional improvement in standardized preclinical models of closed head injury [101–103], as well as preclinical models that were designed to mimic the heterogenous pathology associated with TBI, including blast injury [73], subarachnoid hemorrhage [103], intraparenchymal hemorrhage [104], and cerebral ischemia [105, 106]. Based on these compelling preclinical data of neuroprotection across the different facets of clinical TBI pathology [92], CN-105 was selected for further clinical development. Initial Phase 1 trials were performed in Western [107] and Asian populations [108] populations. These phase 1 studies confirmed a strong safety profile after intravenous dosing, a terminal half-life of approximately 3.5 h, and no significant drug accumulation upon repeated dosing [107, 108].

Fig. 2.

Figure demonstrating the amino acid structure of the receptor binding face of ApoE with the five peptide residues that comprise CN-105 highlighted in blue. Ala alanine, Arg arginine, Asp Aspartate, His histidine, Leu leucine, Lys lysine, Ser serine, and Val valine (used with permission: Guptill JT, Raja S, Ramey S, et al. Phase I randomized, double-blind, placebo-controlled study to determine the safety, tolerability, and pharmacokinetics of a single escalating dose and repeated doses of CN-105 in healthy adult subjects. J Clin Pharmacol. 2017;57(6):770–776)

Clinical Development of CN-105

Because CN-105 appears to affect maladaptive neuroinflammatory and excitotoxic pathways following acute brain injury, there is a spectrum of neurological indications for which it might be effective in addition to TBI. For example, as noted above, the preclinical efficacy of this approach has been demonstrated in preclinical models of stroke, blast injury, subarachnoid, and intraparenchymal hemorrhage, all of which are characterized by secondary tissue injury mediated by glial activation and neuroinflammation. Important criteria for the first in-disease state trial of CN-105 included relevance to TBI pathology and tractability of pilot trial design. Based on these considerations, intracranial hemorrhage was chosen as the initial therapeutic outlet, as parenchymal hemorrhage is a common component of TBI, and primary hypertensive intracranial hemorrhage (ICH) tends to have a more focal and homogenous pathology, which allows for the longitudinal radiographic monitoring of cerebral edema as a marker of target engagement.

The resulting trial, CN-105 in participants with acute supratentorial intracerebral hemorrhage (CATCH; NCT03168581) incorporated a dose of 1 mg/kg CN-105, administered intravenously at 6-h intervals for 72 h [109]. CATCH was designed as a multi-site, open-label study to establish molecular and radiographic markers of target engagement and surrogate markers of efficacy. Although primarily designed for safety and feasibility, to establish a preliminary signal of efficacy, CN-105-treated participants were compared 1:1 with participants closely matched for sex, race, and presentation ICH score that were drawn from the Ethnic/Racial Variations of Intracerebral Hemorrhage (ERICH) study, a cohort of 3000 patients with ICH prospectively assembled between 2010 and 2015 [110]. The results were promising, as CN-105-treated patients demonstrated improved 30-day functional outcomes (lower modified Rankin score) when compared to matched controls (odds ratio 2.69; 1.31–5.51 95% CI) [109]. Based on these encouraging results, we recently initiated a multicenter, randomized, double-blind, placebo-controlled clinical trial for patients with acute intracerebral hemorrhage in Singapore (S-CATCH; NCT03711903) [111].

In addition to these ongoing studies, the anti-inflammatory and neuroprotective effects of CN-105 are also currently being evaluated in the setting of postoperative cognitive dysfunction (POCD). Delirium and cognitive dysfunction following surgery represents a significant public health issue, especially in patients over the age of 60, and is associated with decreased quality of life and 1-year mortality [112]. Although the mechanisms underlying POCD are likely multifactorial, increasing evidence implicates maladaptive neuroinflammation leading to blood–brain barrier breakdown and secondary tissue injury [113]. To study the effects of CN-105 in this setting, the modulating ApoE signaling to reduce brain inflammation, delirium, and postoperative cognitive dysfunction (MARBLE; NCT 03802396) study was designed as a single-center, randomized, tiered, dose-escalation study evaluating perioperative neurocognitive disorders in patients undergoing prolonged non-cardiac surgery [114]. Of note, this study is also designed to provide information on the effect of CN-105 on cerebrospinal fluid markers of injury and inflammation and their correlation with neurocognition, which may be of relevance in evaluating the delayed effects of TBI.

CN-105 as a Potential Therapy in TBI

Given the societal impact of traumatic brain injury, the development of a pharmacological intervention to improve functional outcome has remained an intense focus of research. However, as noted above, despite hundreds of clinical trials the identification of successful therapeutic strategies has remained elusive. One major obstacle to the development of an effective pharmacological intervention is the heterogeneity of tissue pathology associated with TBI, which may include elements of subdural, epidural, intraparenchymal, and subarachnoid hemorrhage as well as cerebral ischemia, tissue contusion and diffuse axonal injury. Thus, it is reasonable that preclinical candidate therapies be chosen based on the demonstration of histological and functional improvement in preclinical models of these different facets of TBI pathology. Improvement in multiple preclinical models has generated much interest in apoE mimetic peptides for TBI and has been one of the major criteria for the selection of CN-105 (Table 1) [90, 96, 97, 101, 102, 115–120].

Table 1.

Preclinical studies evaluating apoE mimetic peptides in TBI

Injury	Species	Peptide/dosing	Histological/biochemical outcome measures	Functional outcome measures	References
TBI (closed head injury)	Mouse	ApoE (133–149) high dose (406 1 µg/kg), low dose (203 µg/kg) or scrambled control IV 30 min post-injury	Reduction in oxidative stress (aconitase), neuronal degeneration, and TNFα mRNA	Improved vestibulomotor function and memory in peptide treated mice	Lynch et al. 2005 [90]
TBI (closed head injury)	Mouse	COG1410 0.6 mg/kg IV 120 min post-injury × 1 dose	Reduction in degenerating neurons, microgliosis	Improved vestibulomotor function and memory	Laskowitz et al. 2007 [96]
TBI (cortical contusion)	Rat	COG1410 0.8 mg/kg IV, 0.4 mg/kg IV 30 min post-injury	Smaller lesion volume and reduction in astrocytosis with 0.8 mg/kg dose	Improved motor outcomes with 0.8 mg/kg dose	Hoane et al. 2007 [115]
TBI (closed head injury)	Mouse	ApoE(133–149) 1 mg/kg IV × 1 vs. 100 µL isotonic saline 30 min post-injury	Reduction in degenerating neurons, microgliosis, TNFα, and Aβ	Improved vestibulomotor function	Wang et al. 2007 [116]
TBI (cortical contusion)	Rat	COG1410 0.8 mg/kg IV at 30 min and 24 h post-injury	Reduction in degenerating neurons	Improved sensorimotor function and working memory	Hoane et al. 2009 [97]
TBI (closed head injury)	Mouse	COG1410 1 mg/kg IV after scalp sutured close post-injury followed by 1 mg/kg IV daily until day prior to sacrifice	Suppressed activation of MMP-9, reduced breakdown of blood–brain barrier, reduced TBI lesion volume, and vasogenic edema	Decreased functional deficits compared with saline-treated TBI animals	Cao et al. 2016 [117]
TBI (closed head injury)	Mouse	CN-105 various doses, routes and timepoints administered prior to injury	Reduced hippocampal microgliosis in treatment group	Improved vestibulomotor function	Van Wyck et al. 2022 [101]
TBI (closed head injury)	Mouse	COG1410 1 mg/kg IV daily	Reduced traumatic axonal injury in pericontusional	N/A	Jiang and Brody 2012 [118]

The heterogeneity of patients presenting with TBI also has implications for patient enrollment in clinical trials. Traditionally, clinical trials have relied on the presentation Glasgow Coma Scale (GCS) assessment of injury severity as a key enrollment criterion. However, in isolation, the presentation of GCS may be misleading. Thus, it would be reasonable for future trials to incorporate more quantitative radiographic and serum biomarkers of tissue injury to select patients with a degree of TBI severity that might mostly benefit from a particular therapy. Several protein biomarkers of neuronal injury and glial activation such as glial fibrillary acidic protein (GFAP), neurofilament light protein (NFL), phosphorylated tau protein (p-tau), and ubiquitin C-terminal hydrolase-L1 (UCHL-1), as well as other markers of inflammation have demonstrated promise in early clinical trials [121–123]. Of note, the recent development of point-of-care platforms that measure GFAP and UCHL-1 at bedside may increase the practicality of stratifying TBI severity prior to enrollment [124, 125]. Moreover, it is important to recognize that secondary injuries may commonly occur during the prehospital period where hypoxia, cerebral hypoperfusion, and expanding hematoma may exacerbate secondary tissue injury. Ideally, agents such as CN-105 should be selected based on safety profile which may ultimately facilitate administration in the prehospital setting.

Once appropriate neuroprotective candidates are defined for clinical translation, clinical trials are often difficult to directly compare due to widely varying trial designs. For example, the timing of intervention is a critical variable and should be based on presumptive mechanisms of action, as traditional neuroprotective agents designed to reduce initial excitotoxic injury might be expected to have a short therapeutic window as compared to interventions that mitigate secondary tissue injury by reducing maladaptive neuroinflammatory responses. Ideally, candidate therapies should be chosen with pleiotropic mechanisms of action or be used in combination to target different components of the injury cascade. In this regard, CN-105 was chosen based on preclinical evidence that it mitigated both glutamate-mediated excitotoxicity as well as neuroinflammatory responses.

In addition to the harmonization of trial design to integrate standardized radiographic and biochemical surrogates of tissue injury, the adoption of consensus endpoints that are functionally relevant and sensitive to intervention would increase the likelihood of trial success and allow more direct comparisons of efficacy between trials. Finally, it is important to recognize that trials in TBI are often extremely resource-intensive, and clinical development may be abandoned in an early trial that demonstrates promising but non-statistically significant results due to financial considerations. Similar proprietary and commercial concerns often make it difficult to design trials that test rational drug combinations targeting different components of the injury cascade.

Ultimately, one mechanism to address many of these limitations in TBI trials and accelerate the process of drug development may be the implementation of an adaptive platform design. A platform study is a randomized study design targeting a specific disease that is performed under a master protocol, which allows for the standardization of inclusion/exclusion criteria that incorporate consensus clinical, biomarker, and radiographic assessments, as well as validated clinical outcome assessments that are sensitive to therapeutic intervention. Platform designs also allow for statistical efficiencies which may optimize the use of resources. For example, the ability of multiple interventions to be evaluated simultaneously against a common control group and the addition of new interventions during the trial make this approach particularly attractive for early-phase clinical trials designed to identify promising candidate therapies for further development. Because adaptive platform trials take advantage of the stream of real-time information acquired during a trial, it also allows for statistical efficiencies with regard to conditional powering, which make it less likely that a potentially promising intervention is abandoned due to lack of statistical significance. This approach is being adopted by the TRACK-TBI network, which will be testing multiple promising TBI therapies in a multi-center, double-blind, placebo-controlled adaptive platform [126, 127].

Conclusions

Our understanding of TBI is rapidly evolving. With the development of novel methods for diagnosing TBI, including biomarker panels and advanced imaging, it is reasonable to expect that the incidence of TBI will continue to increase as we improve our diagnostic capabilities. This trend hastens the need for effective neuroprotectant strategies. As studies aiming to develop effective neuroprotectants continue, it will be necessary to consider and address shortcomings in existing trial designs and develop clearly defined and functionally relevant clinical outcomes when evaluating neuroprotectants. The disappointing results thus far of reductionist approaches may support increased efforts to identify endogenous proteins with pleiotropic effects that can affect TBI outcomes. Apolipoprotein E is one of the most studied genetic polymorphisms as it relates to outcomes following neurotrauma and has an adaptive role after TBI through the reduction of neuroinflammation, excitotoxicity, and secondary brain injury. Although the intact apoE lipoprotein is too large to cross the blood–brain barrier, apoE peptides such as CN-105 appear to effectively interact with receptors that can modulate neuroinflammatory and excitotoxic responses to brain injury. Furthermore, CN-105 is stable, demonstrates predictable pharmacokinetics, and has a favorable toxicity profile. A growing body of literature supports its efficacy in a wide variety of head injury subtypes, making it an attractive candidate for continued clinical evaluation in human trials of TBI.

Data Availability

All data supporting the findings of this study are available/cited within the paper. All pre-clinical data has been made public in the referenced works.

Declarations

Conflict of Interest

None.