Breaking Barriers: Characterization of the Intradermal Lipopolysaccharide Challenge as an In Vivo Model for Controlled Induction of Vascular Leakage in Healthy Volunteers

Abstract

Vascular leakage and its associated phenomena vasodilation and endothelial activation are pathophysiological features of various diseases. Multiple drug candidates targeting these phenomena are in development, necessitating translational models to demonstrate proof-of-pharmacology and proof-of-mechanism in early-phase clinical trials. This single-center experimental study evaluated the intradermal lipopolysaccharide (id LPS) challenge model as a tool to induce and characterize vascular leakage in healthy participants. Eight participants (male:female = 4:4) received id LPS in the volar forearms, followed by serial pharmacodynamic assessments, including imaging and suction blister induction up to 9 hours after injection. Id LPS administration resulted in significant increases in skin perfusion (P < 0.0001), erythema (P = 0.0013), and skin volume (P = 0.0008), indicating initial stages of inflammation and fluid extravasation. Blister fluid analysis revealed elevated extravascular concentrations of albumin (P = 0.0011), total protein (P < 0.0001), and neutrophils (P = 0.0342), supporting the presence of vascular leakage. Moreover, the expression of endothelial activation markers VCAM-1 (P = 0.0015), ICAM-1 (P = 0.0004), ITGB1 (P = 0.01), and E-selectin (P = 0.0218) increased significantly. Disruption of endothelial cell-cell integrity was supported by increased expression of VE-cadherin (P = 0.0002) in blister fluid. These findings support the applicability of the id LPS model for the induction of vascular leakage in humans. This model holds potential as a translational tool for evaluating the pharmacodynamic responses of vascular leakage-targeting drugs in early clinical development.

Figure 1

Schematic representation of the study design and schedule of assessments. In this experimental clinical study, eight healthy volunteers received a total of three id LPS injections to designated areas on the volar forearms. One id LPS‐injected site on the arm served as area for non‐invasive imaging prior to LPS injection (baseline) and repeatedly from 0.25 up to 9 hours thereafter. Imaging consisted of recording skin perfusion using laser speckle contrast imaging and capturing erythema and skin volume using multispectral imaging. For albumin, total protein and Olink protein analysis, blister fluid was collected from suction blisters induced at all areas on the arms at baseline and at 0.5, 4, and 9 hours after id LPS injection. Blood samples for albumin and total protein analysis were taken simultaneously with blister induction. h, hour; Id, intradermal; LPS, lipopolysaccharide.

Figure 2

Non‐invasive quantification of local dermal responses to id LPS to assess vasodilation and vascular leakage. Skin perfusion was non‐invasively assessed using laser speckle contrast imaging, and erythema, and skin volume were evaluated using the Antera 3D camera. (a–c) Representative clinical imaging data from a participant demonstrating the observed local response in (a) skin perfusion, (b) erythema, and (c) skin volume at 0.5 h, 4 h, and 9 h after id LPS injection as compared to baseline. (d–f) Group summary data (n = 8) showing mean ± SD change from baseline in (d) skin perfusion, (e) erythema, and (f) skin volume over time after id LPS injection. AU, arbitrary unit; CFB, change from baseline; h, hours; LPS, lipopolysaccharide.

Figure 3

id LPS‐induced increases of protein contents in local blister fluid as a measure of vascular leakage. (a) Concentrations of albumin and total protein (g/L, mean ± SD) in blister fluid at baseline prior to LPS injection, and at three LPS‐injected skin areas at 0.5, 4, and 9 hours after id LPS injection. (b) Corresponding serum albumin and total protein concentrations (g/L, mean ± SD) measured from blood samples collected in parallel with the induction of the suction blisters. LPS, lipopolysaccharide.

Figure 4

id LPS‐induced alterations in endothelial activation and endothelial junction markers in local blister fluid. (a) NPX values representing the expression of proteins associated with endothelial activation (VCAM‐1, ICAM‐1, and E‐selection) in blister fluid at baseline and at 0.5, 4, and 9 hours after id LPS injection. (b) NPX values for proteins associated with endothelial cell–cell interaction (VE‐cadherin and endothelial surface adhesion marker (ESAM)) in blister fluid collected from skin suction blisters on a designated control area (baseline) and on three LPS‐injected sites (0.5, 4, and 9 hours after injection) on the volar forearms. (c) Quantification (individual values and mean ± SD) of neutrophil infiltration at the designated skin area at baseline prior to LPS injection, and at three LPS‐injected skin areas at 0.5, 4, and 9 hours after id LPS injection. Neutrophils were measured as the absolute number of CD45+/CD66b+/Siglec8− cells in 100 μL blister fluid using flow cytometry. Each boxplot in (a) and (b) represents the median, interquartile range, and 1.5*IQR (whiskers) NPX for each protein. Asterisks indicate levels of statistical significance as analyzed using the Eisinga, Heskes, Pelzer & Te Grotenhuis post hoc test. *P < 0.05, **P < 0.01, ***P < 0.001. NPX, normalized protein expression.

Figure 5

Schematic representation of the effects of an intradermal LPS challenge on vascular leakage and potential read‐outs. This figure displays a cross section of part of the dermis, the vascular wall and part of the vascular lumen. (a) The inflammatory challenge is initiated by the intradermal injection of LPS into the skin of healthy volunteers. (b) The id LPS injection elicits an innate inflammatory response characterized by vasodilation (increased perfusion and erythema), leukocyte attraction, and endothelial activation. (c) Furthermore, the endothelium is activated (e.g., VCAM‐1 and ICAM‐1), leading to extravasation of neutrophils, and vascular integrity is altered by disintegration of tight junctions and the extravasation of fluids and proteins (e.g., albumin). Non‐invasive available experimental read‐outs for the described processes are imaging of vasodilation by laser speckle contrast imaging, and erythema and skin volume as measure of fluid extravasation by Antera 3D camera. Invasive experimental read‐outs are analysis of local suction blister fluid by laboratory assays for albumin and total protein, Olink proteomics for endothelial markers, and flow cytometry for neutrophil extravasation. ICAM‐1, intercellular adhesion molecule 1; LPS, lipopolysaccharide; VCAM‐1, vascular cell adhesion molecule 1.