Engineering Intravenously Administered Nanoparticles to Reduce Infusion Reaction and Stop Bleeding in a Large Animal Model of Trauma

Abstract

Bleeding from traumatic injury is the leading cause of death for young people across the world, but interventions are lacking. While many agents have shown promise in small animal models, translating the work to large animal models has been exceptionally difficult in great part because of infusion-associated complement activation to nanomaterials that leads to cardiopulmonary complications. Unfortunately, this reaction is seen in at least 10% of the population. We developed intravenously infusible hemostatic nanoparticles that were effective in stopping bleeding and improving survival in rodent models of trauma. To translate this work, we developed a porcine liver injury model. Infusion of the first generation of hemostatic nanoparticles and controls 5 min after injury led to massive vasodilation and exsanguination even at extremely low doses. In naïve animals, the physiological changes were consistent with a complement-associated infusion reaction. By tailoring the zeta potential, we were able to engineer a second generation of hemostatic nanoparticles and controls that did not exhibit the complement response at low and moderate doses but did at the highest doses. These second-generation nanoparticles led to cessation of bleeding within 10 min of administration even though some signs of vasodilation were still seen. While the complement response is still a challenge, this work is extremely encouraging in that it demonstrates that when the infusion-associated complement response is managed, hemostatic nanoparticles are capable of rapidly stopping bleeding in a large animal model of trauma.

Figure 1:

Characteristics of nanoparticles used in this study. (A) Schematic of the nanoparticles. All of the nanoparticles consisted of a polyester degradable core (either PLGA-PLL or PLA) with PEG arms (Mn~4600 Da). The hemostatic nanoparticles all had an RGD peptide (either GRGDS or cRGD. The control nanoparticles in the study all had no peptide. (B) Representative scanning electron micrograph of hemostatic nanoparticles. DLS confirmed that the size was approximately 445+/−102 nm for all of the particles used in the study unless otherwise noted. (C) The peptide density was determined as a function of peptide per lysine unit and was 0.3 peptides per lysine.

Figure 2:

(A) Timeline of the injury and treatment. (B) Blood loss over 60 minutes with saline infusions at 15, 30 and 60 minutes. The dotted lines represent the SE for each timepoint (n=4). (C) Schematic of the injury. The left lobe (LL) is isolated from the underlying anatomy and medial lobe (left LML; right RML) with a malleable retractor and measured and marked with cautery 2” from the apex (1). Two additional measurements are made from the apex to the lateral aspects of the resection line to ensure consistent equilateral angles (2 & 3). Ring clamps are used to hold the liver while the injury is made. The liver is resected to the left lobe midline (1), starting from patient---right. This is allowed to bleed for 1 minute with ring clamps still holding proximal to the injury line, and then the remaining liver is cut. After the injury is made, the left lobe is placed back in its natural resting place to prevent alteration of normal hepatic blood flow. VC=hepatic inferior vena cava. (D) Cross section of a resected section of the liver lobe showing the major vessels. (E) Quantification of vessel diameters at surface of resected liver

Figure 3:

Injury at time=0. Particles were administered at time= 5min. (A) 2 mg/kg of hemostatic nanoparticles or control nanoparticles triggered vasodilation and bleed out within minutes of administration leading to death. (B) Table summarizing survival time and total blood loss for the first-generation nanoparticle group at 2 mg/kg. In contrast, the saline group (n=4) survived for the entire experiment with an average blood loss of 722+/−106 ml.

Figure 4:

No injury was performed. (A and B) 2 mg/kg PLA-PEG Nanoparticles (Zeta potential = −30 mV) were administered at time=0 minutes. Within 2 minutes the heartrate and blood pressure changed dramatically followed by a subsequent spike at t=8–12 minutes. (B) The blood gases also showed the same dramatic changes over the same time period. This is consistent with what has been seen by others following nanoparticle administration (48, 63, 64). (C and D) 2 mg/kg of PLGA-PLL-PEG Nanoparticles (Zeta = 22.97 mV) were administered at time=68 minutes. These positively charged particles led to a similar response as in A and B. (E and F) 2 mg/kg of PLGA-PLL-PEG Nanoparticles, (Zeta = 1.29 mV) were administered at time=0 minutes and did not show signs of these physiological responses.

Figure 5:

Second generation hemostatic nanoparticles. (A) At the lowest dose, 0.8 mg/kg, all of the zeta potentials for nanoparticles did not increase the bleeding rate following administration. The zeta potentials in (A) represent the particles we designated as neutral with −3mV

Figure 6:

The impact of the second-generation hemostatic nanoparticle (hNPs*) and control nanoparticles (cNPs*) on bleeding over the first 60 minutes following the liver injury model. The particles were delivered five minutes after the injury was made (at time=5 minutes in each graph). All the nanoparticles included in this data had reproducible zeta potentials between −3 and 3 mV with small standard deviations. (A) 0.8 mg/kg dose (B) 2 mg/kg dose (C) 3.3 mg/kg dose

Figure 7:

(A) Complement response as measured by changes in C3 and C3a between baseline (BL) and 15 minutes post injury. These data would suggest that the particles do not activate complement, yet in a subset, marked by the line, we see signs of the complement associated infusion reaction marked by vasodilation and small changes in the physiological parameters. (B) A cytokine array panel on serum from baseline and 15 minutes post injury in an infusion suggests that the impact of particle may be better understood through looking at different molecules. The cNPs* were given at a 0.2 mg/kg dose and the presence of PVA was, ultimately, identified as a trigger for the infusion-associated complement reaction.