One of the first groups that considered this potential application for injectable particles was Coller et al. They looked at the use of what they called ‘thromboerythrocytes.’¹⁰ These particles use a peptide containing the sequence arginine-glycine-aspartic acid to bind to activated platelets, which display the GPIIbIIIa receptor. This sequence was bound through a heterobifunctional crosslinking agent to the surface amino groups on erythrocytes. In prior work, they had shown how important the length of the crosslinker is for establishing interaction between the activated platelets and the RGD linked particles³⁶ They were then able to make thromboerythrocytes that interacted selectively in vitro with platelets activated by the presence of ADP. This work established the importance of the RGD sequence in development of particles capable of interacting with activated platelets.

The next group that looked at an intravenously injectable particle to improve clotting was Levi et al. They developed particles made of albumin coated with fibrinogen (a blood protein that contains the RGD sequence).¹¹ These particles were approximately 3.5–4.5 microns in size, larger than many of current particle formulations. They showed a reduction in bleed time in an in vivo ear injury model in thrombocytopenic rabbit. In this model they pre-injected the rabbits with the particles prior to injury. They did show with this model that the particles significantly reduce the bleed times of the animals, demonstrating that particles using the RGD sequence to bind activated platelets can, indeed, have an effect on bleeding outcomes.¹¹ However, the size of these particles is problematic because particles of that size can be filtered in the capillary beds of the lungs.³⁷

Another approach to the generation of particles that can perform this function is the use of modified platelets, which can both initiate and strengthen clotting. This is the method that Fitzpatrick et al. have looked at. They have shown that they can collect and freeze-dry platelets stabilized with trehalose in a form that is still similar to native platelets.¹² The majority of the particles are in the 1–5 micron range (the size range of native platelets). In vitro studies show that the particles display GPIIbIIIa and can still perform the functions necessary for hemostasis. In vivo studies show that the particles stay in circulation after injection for the same amount of time as a normal platelet transfusion and no adverse effects of injection were apparent. In vivo testing with the ear bleed model in thrombocytopenic rabbits showed that the particles also significantly reduced blood loss.^12,13 Safety testing of these particles was also done in a non-human primate model (rhesus macaque) without any adverse events reported. However, biologically derived particles may also lead to potential for adverse immune responses and disease transfer.³⁸

Tranexamic acid is a strong antagonist of plasminogen activation, and therefore acts as an antifibrinolytic. There is convincing evidence that its administration after trauma greatly improves outcomes, reducing (all cause) mortality by 23.9%-17.4%, in one study of 896 wounded soldiers at a military hospital in Afghanistan.^14,15 However, it has the potential drawbacks of increasing risk of diffuse intravascular coagulation, increased risk when administered 3–8 hours after trauma, and increased risk when co-administered with blood products. The concurrent use of a permissive hypotension approach may also increase tranexamic acid’s risks since this approach leads to a decreased glomerular filtration rate, and hence a delayed clearance rate. This may lead to a change in the drug’s efficacy and safety.¹⁴ Nevertheless, it has tremendous potential, and may become incorporated in standard damage control resuscitation.³⁹

The administration of recombinant factor VIIa intravenously to reduce bleeding after acute trauma has been a topic of debate.^5,19,20,40 Several studies have shown that perioperative administration of rFVIIa reduces the volume of blood transfusion. However, it is unclear whether the benefit is large enough to have any associated effect on mortality after hemorrhagic trauma.^16–18 Its potential use in the prehospital phase is further diminished due to its high cost, potential for adverse effects, and necessity to be stored at 2–8 degrees C.^18,20

Liposomes are a popular type of synthetic nanoparticle used primarily for drug delivery. Okamura et al. have developed hemostatic liposomes coated with an alternative peptide: the dodecapeptide HHLGGAKQAGDV (H12) which was shown in vitro to suppress platelet aggregation, suggesting interaction with activated platelets. The liposomes with this peptide are much smaller than the comparable particles derived from cells: they are typically around 260nm in diameter (+/− 60 nm). In addition, liposomes allow encapsulation of drugs, in this case ADP within area enclosed by the lipid bilayer. The use of ADP there may enhance platelet aggregation by activation of additional platelets.^22,23 These particles are entirely synthetic, but still offer a significant improvement in bleeding time in both thrombocytopenic rat and thrombocytopenic rabbit models when compared with saline as a treatment.²²

Bertram et al. have also shown that synthetic particles can be a functional alternative to particles derived from blood cells and proteins. These particles are made with a biodegradable polymer core (poly(lactic-co-glycolic acid) with polyethylene glycol as the spacer used to attach the RGD peptide. They have been shown to be effective in a rat femoral artery injury model at significantly reducing bleeding time.²⁴ These particles were also later shown to effectively reduce lethality in a rat liver injury model by Shoffstall et al.²⁵ The trend toward synthetically based particles offers several advantages. Size can be more easily controlled with synthetic particles and tend to be smaller, which may be an advantage. In addition, both the liposomes and the particles with a polymer core can be drug loaded to offer additional treatment at the injury site if desired.

The pig is the standard model for uncontrolled hemorrhagic trauma, when investigating the physiological impact of a potential therapy.^{62–65,73–76} The cardiovascular system is well-correlated with human parameters and the comparable size allows for devices to be used in both clinical and research environment without modification.⁴¹ Furthermore, the wound-healing process appears to be similar to the human due to similarities between porcine and human skin.⁴¹ However, a porcine model is more expensive due to the equipment and need to for a technically-trained staff.⁴¹ Furthermore, there is an increasing body of evidence that suggests pigs may be especially sensitive to complement activation and related pseudoallergy (CARPA).^77–79 Briefly, a pseudoallergy is observed after intravenous infusions of certain nanoparticle formulations. Symptoms include severe hypotension, cardiopulmonary dysfunction, and, if severe enough, death. Interestingly, other species are less susceptible to symptoms. However, it has been observed directly in both swine and dogs.^79,80

Non-human primates are the best animal analog of humans available. However, this makes their use in medical research controversial and, due to the administrative overhead, expensive. Thus, the use of primates in trauma is not well suited for establishing efficacy in experimental therapies, except where the anatomical/physiological differences in other animals precludes their use. The majority of non-human primate studies (in hemorrhagic trauma) are for preclinical studies, where the main goal is testing safety.

Several hemorrhagic trauma models have been established.^12,81–83 Recently, of particular interest, Thrombosomes (a lyophilized hemostatic agent derived from platelets), underwent a safety study in rhesus macaques, in the presence of a hemorrhagic liver trauma.¹²

The adsorption of proteins is also a significant concern for any surface or particle that will contact blood because it is believed to cause rapid uptake of particles by the RES system. It is well understood that when the surface of a biomaterial is put into the body, it is immediately covered with a protein corona.^85,86 Understanding the degree to which nanoparticles will be taken up by phagocytosis and maximizing evasion of the RES system are considered necessary for the success of nanoparticle treatments. In addition, the formation of the protein corona alters the interaction of nanoparticles with their environments, which may effect the ability of nanoparticles targeted to activated platelets to interact with their receptors.^85,86 PEGylation of nanoparticles is the most typical method used to attempt to overcome protein adsorption. It has been shown that for polymeric nanoparticles, PEG content between 2 and 5 weight %, with a PEG molecular weight of 5000 grams/mole is optimal for prevention of protein adsorption.⁸⁷ Even with the use of PEG to prevent protein adsorption, there will still be protein adsorption, and the result is that these types of nanoparticles will have a relatively short circulation time.

Particle stability in storage is another concern for successfully translating particles to a clinical setting. The need for refrigeration significantly hurts the chances that a technology will be successful in a military setting or useful for first responders. Alternative treatments, such as blood transfusions and rVIIa require refrigeration, which is a significant impediment to use in the field. Some of the proposed treatment can be stored at ambient temperatures, such as the thrombosomes.¹³ Storage of liposomes has proven challenging. They are most easily stored in suspension, which is stable for several months when refrigerated.⁸⁸ They can be stored longer by freeze drying, but will still suffer degradation, and keep best when maintained at cold temperatures, although they do have a glass transition in the range of up to 40 °C, beyond which no particle structure would be expected to be maintained.^89,90

Polymers offer a potentially more stable solution. Polymeric nanoparticles are most often freeze-dried, which is more stable than storage in aqueous solution.⁸⁸ While PLGA, used by Bertam and Shoffstall et al., has a relatively low glass transition temperature of around 40 °C, PLA, another polymer used frequently for nanoparticle formulation, is an alternative polymer that has a higher glass transition of 50 °C, which could offer a solution to temperature degradation. It is also crystalline, which offers additional stability, and will form stereocomplexes when both L- and D- lactide polymers are present.^91,92

Nanoparticles are stored in suspension, frozen, or as a lyophilized product, depending on their stability in each phase. For biodegradable polymer nanoparticles, they are generally lyophilized for long-term storage. However, due to crystallization during processing, aggregates are formed, which then require sonication in order to recover pre-lyophilization size distributions.⁹³ This has led to the development and use of cryo- and lyo-protectants. Some of these involve sugars such as trehalose and sucrose, polymers PVA, poly(vinyl pyrrolidone) (PVP), and even matrices, such as gelatin.⁸⁸ Some concerns exist as to the concentrations of these excipients required in order to stabilize nanoparticles, for example a 1:1 w/w ratio in the case of trehalose, which could constitute a substantial dose of sugar depending on the application.⁸⁸ While trehalose has been shown to be safe at oral doses up to 50g, its metabolism with i.v. delivery may require further investigation in regards to safety and toxicity.⁹⁴ The use of surfactants such as Tween 80 (polysorbate 80), and other excipients are used to minimize the energy required to resuspend nanoparticles after lyophliiziation, however toxicity is always a potential concern.⁹⁵

Complement activation related pseudoallergy (CARPA) is elicited readily in pigs during administration of certain liposomes and several polymeric-based nanoparticles.⁷⁷ The hallmark symptoms of CARPA appear within 1–3 minutes following particle administration and include cardiopulmonary distress (including increase in heart rate, hypotension, decreased cardiac output, decreased pulmonary pressures, and decreased blood gas levels), and a characteristic flushing of the skin (erythema) upon reperfusion of the tissues.^77–79 Surprisingly, unless symptoms are so severe they lead to mortality, these issues spontaneously resolve within minutes. Regardless, the consequences of a severe CARPA episode during administration of intravenous hemostats could be catastrophic, and the mitigation by pharmaceutical prophylaxis is likely contraindicated. Therefore, the development of robust models to ensure a minimized risk for CARPA during nanoparticle administration, and elucidating its mechanisms so that patient-exclusion criteria, based on heighted risk factors, will be paramount to the clinical translation of intravenous hemostatic technologies.

Complement activation is classically induced by traumas, invading bacteria, and viruses; experimentally, it is induced with endotoxin lipopolysaccharide (LPS). Pigs, compared to humans, are especially sensitive to CARPA, endotoxemia and sepsis, which make them good models for those pathologies.⁹⁶ Administration of Doxil (a liposome-based chemotherapeutic) causes mild-to-moderate hypersensitivity reactions in 0–25% of humans⁹⁷, sometimes even subclinically, yet causes moderate-to-lethal reactions in swine.⁹⁸ Interestingly, drastic response to nanoparticles in mice and rat models are not readily observed, making them very poor models for detecting CARPA. It is most likely, as has been previously reported, that the pig species simply has a heightened immune response to pathogens, potentially due to their selective breeding in the meat industry, where susceptibility to disease has been strongly selected against.⁹⁶

While these reactions are thought to be pseudoallergies, in at least one study, repeat administrations (of a novel pegylated micelle) increased the severity of reactions, measured by histamine release and complement activation, consistent with antibody-mediated reactions.^99,100 Since the precise mechanism has not yet been elucidated, methods to design around this adverse reaction are also somewhat poorly understood. Three main strategies have been employed to date: prophylaxis, tachyphylaxis, and tuning surface chemistry.