Monocyte depletion reduces late experimental post-thrombotic fibrotic injury in a stasis mouse model

Abstract

Background:

Post thrombotic syndrome is a fibrotic disease related to inflammation resolution. There are no direct therapies that can ameliorate this disease. Monocyte/macrophages (Mo/MØ) are the primary leukocyte involved with later venous resolution, and likely direct vein wall responses and healing. Herein, we explored the vein wall response with Mo/MØ depletion by two methods.

Methods:

Using two mouse models of venous thrombosis (VT), complete stasis and a flow restricted model, Mo/MØ depletion was accomplished using CD11b-DTR mice administered diphtheria toxin, and clodronate micelle administration in wild type mice. Tissue assays for structural histology, immunohistochemistry and western blotting were performed.

Results:

Mo/MØ depletion resulted in significantly less vein wall fibrotic thickness, in the stasis model at day 14 in both the CD11b-DTR mice and those receiving clodronate micelles. No significant effect of Mo/MØ depletion was observed in the flow restricted VT model. The decrease in vein wall fibrosis was associated with fewer DDR2+ vein wall cells in the CD11b-DTR mice, but no difference in endothelial luminal coverage. Decreased cytokine and growth factor expression of IL-6, FSP-1, and VEGFa were associated with Mo/MØ depletion. Lastly, PMN depletion was associated with increased proinflammatory Mo/MØ, and a trend towards increased vein wall fibrosis as compared with controls.

Conclusion:

Mo/MØ direct the post VT late fibrotic response, possibly by affecting fibroblasts and inflammatory cytokine expression. This effect was only found with the complete stasis model and is consistent with worsened PTS in humans with complete venous obstruction.

1. Introduction

Post-thrombotic syndrome (PTS) is a late sequelae of DVT that affects 6 to 7 million patients [1] and between 400,000 and 500,000 patients have skin ulcerations [2]. While impaired thrombus resolution in humans correlates with PTS [3], no therapies directly address this entity. Invasive approaches to clear DVT include pharmaco-mechanical thrombolysis, but the recent ATTRACT trial failed to produce paradigm changing therapies for PTS [4], including no significant benefit as compared with anticoagulant therapy. Thus, new approaches remain needed to reduce PTS [5].

Venous thrombosis (VT) is an inflammatory process. This process resembles sterile wound healing [6], with phases of neutrophil (PMN) and monocyte/macrophage (Mo/MØ) influx, followed by fibrosis. Consequently, the post-thrombotic vein walls become stiff, and thicken further with interstitial matrix and myofibroblast accumulation [7]. Experimental DVT models suggest the vein wall injury is dependent on the mechanism of thrombogenesis (stasis or non-stasis), the duration of thrombus-vein wall contact, leukocyte infiltration, and multiple other factors [8]. The histological profile of the late post thrombotic mouse vein is similar to humans [9].

The specific role of the Mo/MØ in vein wall injury and VT resolution is only partially understood, but involves factors such as urokinase plasminogen activator, proinflammatory cytokine and growth factor release, and matrix metalloproteinase-2/9 activity [10,11]. Monocyte/ macrophage polarization is complex but driven in part by environmental cytokines, phagocytosis, and epigenetic mechanisms [10,12]. Herein, we investigate the role of Mo/MØ in modulating late vein wall injury using a conditional monocyte depletion model, as well as exogenous monocyte depletion. We found that monocyte depletion was associated with vein walls characterized by less fibrosis and fewer myofibroblasts and was dependent on the model of thrombosis.

2. Methods

2.1. Animals

Male mice (C57BL/6J, 20–30 g, wild types [WT]) were used for all studies and were purchased from Jackson Laboratory (Stock #000664, Bar Harbor, ME, USA). Male mice were used, as their vasculature is less complex than their female counterparts which lessens the risk of bleeding during surgery and less risk of ovarian necrosis [13]. CD11b-DTR mice were purchased from Jackson Laboratories (stock #006000). They were bred on a C57BL/6J background to be homozygous in the human integrin alpha M (ITGAM) promoter (CD11b), permitting transient depletion of macrophages in various tissues following diphtheria toxin administration (DTx) [14]. For all procedures, mice underwent general anesthesia with 2.0–2.5% isoflurane and oxygen gas with continuous monitoring on a 37 ^◦C heating mat. All animal studies were performed with University of Michigan’s Institutional Animal Care and Use Committee approval.

2.2. VT mouse models

VT was formed via generation of stasis by infrarenal inferior vena cava (IVC) ligation [13]. Briefly, mice were anesthetized and a midline laparotomy was performed. Venous side branches and dorsal branches were interrupted and the infrarenal IVC was fully ligated with a 7–0 polypropylene suture to facilitate stasis thrombosis. To generate a stenosis VT with partial blood flow, after dissection of the IVC, a 30-gauge needle was placed next to the IVC, the ligation was tied around both the needle and the IVC, and the needle was removed. The abdominal wall was secured with 5–0 polyglactin 910 suture. The skin incision was then secured with skin glue or 7 mm Reflex wound clips (Roboz Surgical, Gathersburg, MD), and mice were recovered under a warming lamp. The IVC and thrombi were harvested at 8- and 14- days post ligation following euthanasia via exsanguination. Prior to processing, the IVC and thrombus were measured (cm) and weighed (grams) en bloc and separately [13].

2.3. Drug treatment methods to induce leukocyte depletion

2.3.1. Diphtheria toxin administration

To deplete MO/MO, diphtheria toxin (Enzo Life Sciences, INC., Farmingdale, NY) was administered on two distinct schedules relative to stasis ligation to facilitate the examination of thrombogenesis and thrombus resolution. In the context of thrombogenesis, mice were administered a 10 ng/g loading dosage of diphtheria toxin 24 h prior to surgery. For the assessment of thrombus resolution, mice were injected with the same dosage 48 h after IVC ligation and subsequently every 48 h thereafter. All treatments were administered via intraperitoneal injection (IP). Control groups were injected in the same fashion with 0.9% sodium chloride [14].

2.3.2. Clodronate and control liposome administration

As a second approach to deplete Mo/MØ, C57BL/6 mice received clodronate liposomes (Liposoma, Amsterdam, The Netherlands) at a dosage of 0.1 mL/10 g through intravenous injections via the penile vein. Injections were administered every other day following stasis ligation, and every 48 h thereafter until their scheduled harvest date. Mice in the control group followed the same schedule as the experimental mice, and they were administered 0.1 mL/10 g of 0.9% sodium chloride via the penile vein to mirror the volume of fluid that experimental mice received.

2.3.3. Neutrophil depletion

C57BL/6 mice were injected with 200 μg InVivoPlus anti-mouse Ly6G clone 1A8 (BP0075–1, Bio X Cell, Lebanon, NH) diluted in InVivoPure pH 7.0 Dilution Buffer (Bio X Cell, Lebanon, NH), IP. An initial dose was administered one day prior to venous thrombosis (VT), followed by one day post-VT and then, every 3 days until harvest. Control groups received InVivoPlus rat igG2a isotype control (BP0089, Bio X Cell, Lebanon, NH) diluted in InVivoPure pH 6.5 Dilution Buffer (Bio X Cell, Lebanon) at the same dose and frequency.

2.4. Histology

Fresh tissue (IVC with thrombus intact) was fixed in 10% neutral buffered formaldehyde for 4–8 h, transferred to 70% ethanol, and subsequently embedded in paraffin for immunohistochemistry (5 μm tissue sections, 3 sections per slide). Antigen retrieval was performed using heat-mediated Sodium Citrate buffer (10 mM NaCl solution, pH 6.0 at 95 degrees for 10 min then allowed to cool for 20 min). A species-specific polymer reagent kit was used for animal serum nonspecific antigen binding site block and secondary application (Impress Polymer Detection Kits, Vector Laboratories, Burlingame, CA). Tissue sections were stained for anti-DDR2 (1:750, Abcam, Cambridge, MA), anti-CCR2 (1:200, Novus Biologicals, Centennial, CO), anti-CD31 (1:4000, Abcam, Cambridge, MA) and anti-Arg-1 (1:1500, NBP1–54621, Novus Biological, Littleton, CO). The polymer secondary reagent was applied followed by DAB eqV Peroxidase Substrate (diaminobenzidine, Vector Laboratories, Burlingame, CA). The slides were counterstained with hematoxylin and cover slipped with Cytoseal 60 (Thermo Scientific Richard-Allan Scientific, Waltham, MA, USA). Cells were quantified in a blinded fashion, with positive cells counted and totaled using a Nikon B400 brightfield microscope with Spot camera in eight high-power fields (magnification, 1000×) radially around the IVC wall [15].

Masson’s Trichrome staining for vein wall thickness measurements was performed. Images were taken with Nikon (as listed above) at 100× and analyzed blindly with ImageJ (NIH) software. Eight measurements of the blue collagen ring of the vein wall were taken radially around the IVC, in three sections per animal, and averaged, avoiding measurements between the IVC and aorta.

2.5. CD31 stained for % endothelial preservation

To determine the percentage of endothelial preservation through vein wall remodeling, contiguous CD31 stained luminal and abluminal lengths were measured by IHC described above. Tiled images of stained slides were taken with a Nikon at 200× and analyzed in blinded fashion. Lengths of stained endothelial cells were measured across the luminal and abluminal sides and compared to the total circumferences using ImageJ (NIH) software [16].

2.6. Western blotting

Protein levels of MCP-1, IL-6, TGFb, αSMA, VEGF1, VECAD, and β-actin were measured by immunoblotting from vein wall and/or thrombus harvested tissue [14]. RIPA buffer (ThermoScientific, Rock-ford, IL, USA) and complete ULTRA Mini Tablets (Roche, Mannheim, Germany) were incorporated to isolate protein from thrombus and IVC segments. Protein concentration of these lysates was determined using the BCA assay (Thermo Fisher). Protein separation was then achieved by electrophoresis using NuPAGE 4–12% Bis Tris gels (Invitrogen, Carlsbad, CA). Proteins were then transferred onto PVDF membranes (Millipore, Billerica, MA, USA) and probed with the indicated primary antibodies: anti-VECAD (@1:500, ab2413, Abcam, Cambridge, MA) and anti-VEGF1(@1:1000, ab32152, Abcam, Cambridge, MA). Bound antibodies were subsequently probed with the indicated secondary antibodies. Immunoreactive bands were visualized with super signal west Pico chemiluminescent substrate (ThermoScientific), and densitometry was performed using ImageLab software (BioRad, Hercules, California, USA). Optical densities were summed and normalized to b-actin (@1:20000, sc-47778, Santa Cruz Biotechnology, Dallas, Tx) values using ImageLab software (BioRad, Hercules, California, USA).

2.7. Flow cytometry

For surface staining, IVC/thrombus tissue isolates were collected directly into cold PBS in a 6 well dish, minced thoroughly, morcellated with 18ga needle on a 5 mL syringe, and filtered through a 70 μm cell filter. The filter was washed twice with PBS, and spun down at 2000g at room temp to collect the cells to stain after supernatant was aspirated off. Peripheral blood (submandibular) was collected and after red cell lysis, and cells were processed for surface staining [14]. Cells were stained with a Fixable LIVE/DEAD viability dye (Molecular Probes by Life Technologies; Ref No. L34959; 1:1000 dilution). FcR-receptors were then blocked with anti-CD16/32 (BioXCell, Cat No. CUS-HB-197, 1:200 dilution) for 10 min. Monoclonal antibodies for surface staining included: Anti-CD3 (Biolegend, Cat No. 100304, 1:400 dilution), Anti-CD19 (Biolegend, Cat No. 115504, 1:400 dilution), Anti-Ter-119 (Biolegend, Cat No. 116204, 1:400 dilution), Anti-NK1.1 (Cat No. 108704, 1:400 dilution), Anti-Ly6G (Biolegend, Cat No. 127604, 1:400 dilution), Anti-CD11b (Biolegend, Cat No. 101230, 1:400 dilution), Anti-Ly6C (Biolegend, Cat No. 128035, 1:400 dilution) and Anti-F4/80 (Biolegend, Cat No. 123121, 1:400 dilution). After surface staining, cells were washed twice in Flow Buffer (PBS, 1% BSA, 1% of 2% sodium azide), and biotinylated antibodies were labeled with streptavidin-fluorophore (Biolegend, Cat No. 405208, 1:1000 dilution). Next, cells were washed and fixed with FluroFix (Biolegend, cat # 422101) for 30 min, and then washed with PBS and resuspended into Leibovitz’s L-15 medium (Gibco). These single-cell suspension samples were acquired on a 6-Laser MoFlo Astrios (Beckman Coulter). FACS was performed with Summit Software (Beckman Coulter), analysis was performed using FlowJo software version 10.0 (Tree Star), and data were compiled using Prism software (GraphPad). All populations were routinely backgated to verify gating and purity.

2.8. Statistical analysis

All data are represented as mean ± SD as indicated. Unpaired Studenťs two tailed t-test with Welch’s correction and one-way ANOVA with Bonferroni post-hoc comparison were used for comparison between the experimental groups and controls as appropriate (Graph Pad, Prism, San Diego, CA, USA). A p-value of <0.05 was assigned significance.

3. Results

To assess whether global depletion of the Mo/MØ lineage would impact the late vein wall response, we harnessed the CD11b-DTR mouse with DTx administration to deplete monocytes [17]. This system produces a consistent decrease in circulating monocytes and inhibits VT resolution observed at 8 days in the full stasis model, and did not impact thrombogenesis measured on day 2 [14]. We initiated depletion of Mo/MØ two days after the VT had formed, so as to prevent effects on thrombus formation.

At 14 days, we found that vein wall thickness, a measure of post thrombotic vein wall injury [7,15], was significantly decreased in mice undergoing Mo/MØ depletion in the stasis VT model (Fig. 1A–C). Interestingly, there was no effect on VT normalized weight (weight/length) at this timepoint (Fig. 1D), and all mice had thrombus present. Correspondingly, we found decreased numbers of cells positive for DDR2 (Fig. 1E–G), a myofibroblast marker, but no significant change in the number of cells positive for CCR2 (Supplemental Fig. 1A), a proinflammatory monocyte marker (Ly6C C^hi), or for Arginase-1 (Supplemental Fig. 1B), a pro-reparative monocyte marker (Ly6C^lo) [10]. A reduction in luminal CD31+ endothelial cell coverage was found in the depleted Mo/MØ group, as compared with controls, although not statistically significant (Fig. 1H). Vein wall cytokine and growth factor analysis in the CD11b-DTR group showed decreased IL-6, FSP-1, and VEGFα protein expression (Fig. 1I–K, respectively). We did not find any differences in MCP-1, TGF-β1, or α-SMA protein levels (Supplemental Fig. 1C–E). In contrast to the 14-day timepoint, we found no effect of Mo/MØ depletion on vein wall fibrotic injury at 8 days in the complete stasis VT model (Supplemental Fig. 1F–H).

Fig. 1.

Vein wall fibrosis, fibroblast markers, and cytokine expression, but not endothelial preservation are affected by diphtheria toxin mediated monocyte depletion in CD11b-DTR mice. All panels are from the CD11b-DTx monocyte depletion model. A. Masson’s trichrome analysis of vein wall thickness (μm). B. Masson’s Trichrome stained thrombus and vein wall from a CD11b-DTR NaCl treated mouse. C. Masson’s Trichrome stained thrombus and vein wall from a CD11b-DTR DTx treated mouse. D. Thrombus weight-to-length ratio. E. Number of DDR2⁺ cells per 8 high-power fields (hpf) in the vein wall. F. DDR2 stained thrombus and vein wall from a CD11b-DTR NaCl treated mouse. G. DDR2 stained thrombus and vein wall from a CD11b-DTR DTx treated mouse. H. Percent re-endothelialization along luminal vein wall. I. Immunoblot analysis of vein wall IL-6 expression. J. Immunoblot analysis of vein wall FSP-1 expression. K. Immunoblot analysis of vein wall VEGFα expression. For all panels, each dot represents a separate mouse. DTx, diphtheria toxin; hpf, high-power field; ns, not significantly different. *P < 0.05, **P < 0.01, ***P < 0.001.

To determine if the vein wall fibrosis after monocyte depletion is dependent on the VT model, we performed the same monocyte depletion strategy with a limited flow stenosis model as described [13] using CD11b-DTR mice and DTx administration at the same intervals. In contrast to the stasis model, we found no significant effect on vein wall injury at 14 days post thrombosis (Supplemental Fig. 1I–K).

As an alternative strategy to deplete Mo/MØ, we administered clodronate liposomes to globally deplete monocytes, and which may affect different Mo/MØ subsets than in mice deleted using the CD11b-DTR paradigm [13,18]. Again using the stasis VT model, there was a significant reduction in post-DVT vein wall thickness at day 14 in clodronate-treated mice as compared with controls, consistent with the CD11b-DTR monocyte depletion strategy (Fig. 2A–C). In contradistinction to the CD11b-DTR based Mo/MØ depletion results, a significant increase in VT burden (weight/length) was found (Fig. 2D), and all samples had thrombus present. There was no difference in DDR2⁺ fibroblasts in the vein wall (Supplemental Fig. 2A). As with the CD11b-DTR mice, no significant difference in the CD31⁺ luminal staining was observed, suggesting no direct effect on endothelial preservation (Fig. 2E). Interestingly, fewer Arginase-1⁺ cells were found with the clodronate liposome-treated mice as compared with controls (Fig. 2F–H), and no significant difference in CCR2+ cells were found (Supplemental Fig. 2B). Vein wall cytokine and growth factor analysis showed decreased IL-6 and FSP-1 protein expression (Fig. 2I–J). We did not find any differences in TGFβ1, MCP-1, or VEGFα protein expression between the groups (Supplemental Fig. 2C–E).

Fig. 2.

Clodronate liposome-mediated monocyte depletion reduces vein wall fibrosis, increases thrombus size, and alters cytokine expression. Anti-Ly6C-mediated neutrophil depletion promotes the accumulation of CCR2⁺ monocytes. Panels A–J: Clodronate liposome macrophage depletion model. Panels K-M: C57 mice treated with IgG control or anti-Ly6G antibody in the 8d stasis model. A. Masson’s trichrome analysis of vein wall thickness (μm). B. Masson’s Trichrome stained thrombus and vein wall from a control liposome-treated mouse. C. Masson’s Trichrome stained thrombus and vein wall from a clodronate liposome-treated mouse. D. Thrombus weight-to-length ratio. E. Percent re-endothelialization along luminal vein wall. F. Number of Arg1⁺ cells per 8 hpf in the vein wall. G. Arg1 stained thrombus and vein wall from a control liposome-treated mouse. H. Arg1 stained thrombus and vein wall from a clodronate liposome-treated mouse. I. Immunoblot analysis of vein wall IL-6 expression. J. Immunoblot analysis of vein wall FSP-1 expression. Neutrophil depletion model. K. Number of CCR2⁺ cells per 8 hpf in vein wall. L. CCR2 stained thrombus and vein wall from C57 mouse treated with IgG control. M. CCR2 stained thrombus and vein wall from C57 mouse treated with anti-Ly6G antibody. For all panels, each dot represents a separate mouse. Clod, clodronate; Lipo, liposome; Tx, treatment; hpf, high-power field; ns, not significantly different. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Lastly, Mo/MØ phenotype and activity is dependent in part on their engulfment of cellular debris and the local inflammatory environment as directed by the earlier influx of PMNs [19], we used anti-Ly6C antibody administration to deplete PMNs [20]. We confirmed the early lack of PMNs at day 2 in the vein wall by flow cytometry (not shown). Interestingly, depletion of PMNs preceded significantly increased CCR2+ proinflammatory Mo/MØ vein wall cells at the 8d timepoint (Fig. 2K-M). We assessed both 8- and 14- day vein wall fibrosis. There was no significant difference in vein wall thickness at the 8-day timepoint (Supplemental Fig. 3A–C), with increased vein wall thickness approaching significance at 14 days (P = 0.11; Supplemental Fig. 3D–F). There was no difference in Arginase 1+ cell staining between either timepoints in response to PMN depletion (Supplemental Fig. 3G–H).

4. Discussion

Post thrombotic vein wall injury is a common sequela of DVT and can induce the post-thrombotic syndrome (PTS), for which there are no direct therapies beyond anticoagulation [2]. The post thrombotic vein wall is a classic post-inflammatory fibrotic disease [6,21], offering the potential to modulate the inflammatory response to reduce vein wall injury, as new therapies remain needed to reduce PTS. [5] Herein, we show that in the mouse, post thrombotic vein wall fibrosis is Mo/MØ-mediated and is dependent on the mode of venous thrombogenesis.

The balance of Mo/MØ in injured tissue is dynamic but typically follows an initial neutrophil influx, followed by an early proinflammatory Ly6C^hi monocyte influx, with later predominance of Ly6C^lo Mo/MØ, as observed in several experimental models. For example, in a murine wound healing model, impaired wound healing with persistence of proinflammatory Ly6C^hi Mo/MØ was observed in diabetic mice, as compared to controls [22]. In a murine model of infarcted myocardium, absence of reparative Ly6C^lo Mo/MØ worsened healing and impaired cardiac function [23]. It is likely that in experimental models, the early Ly6C^hi proinflammatory Mo/MØ convert to Ly6C^lo Mo/MØ and allow resolution of inflammation [19,24], although others have shown a circulating Ly6C^lo population invade post inflammatory tissues [25]. In the current study, we did not determine whether Ly6C^hi to Ly6C^lo phenotype conversion occurred or a later wave of Ly6C^lo Mo/MØ influx occurs. Nonetheless, the depletion of monocytes seem to prevent later vein wall fibrotic injury, and we speculate the fibrotic injury is lessened with fewer infiltrating proinflammatory Mo/MØ, and supported in a study focusing on VT resolution [26].

We also showed the post thrombotic vein wall fibrotic injury is thrombosis model dependent, with no effect of Mo/MØ depletion using the stenosis flow-restricted (partial stasis) model. The direct effect of the thrombosis model on vein wall fibrotic injury has been observed in other settings previously [27,28]. This finding is consistent with the human clinical situation, with PTS being more common in those with a totally occlusive DVT as compared to those with less complete or more rapid thrombus clearance [2,3]. We speculate the proinflammatory Mo/MØ promote vein wall injury in the setting of an occlusive thrombus in direct juxtaposition with the vein wall, in contrast to a lesser inflammatory effect with peri-thrombus blood flow. Conversely, restoration of peri-thrombus blood flow is associated with decreased vein wall fibrosis in a mouse model [29], but only with less than 4 days of thrombotic occlusion. We acknowledge a limitation of our study is we did not use clodronate to deplete monocytes in the stenosis model, primarily due to past observations with a greater vein wall injury with stasis as compared with non stasis thrombogenic mechanisms [10,30].

A strength of this study was the use of two techniques to deplete monocytes; the CD11b-DTR mice using DTX, and administration of clodronate liposomes. Interestingly, clodronate may affect PMN functions [31], and may not deplete Mo/MØ as widely as the CD11b-DTR model [18]. We found essentially similar vein wall fibrosis reduction with both depletion strategies, suggesting the Mo/MØ phenotype depletion and injury mechanisms were similar. However, impaired VT resolution with clodronate administration, but not in the CD11b-DTR mice at 14 days, was observed. This suggests that more limited Mo/MØ depletion affects VT resolution. We found fewer pro-reparative Mo/MØ (Arg1+cells) in the vein wall with the clodronate depletion, and this monocyte phenotype is associated with VT resolution [14,26]. Interestingly, a prior study suggested Arginase 1+ cells were correlated with increased late vein wall fibrosis, and is consistent with reduction in these cells observed here [10]. Another mechanism may be that clodronate liposomes have been shown to stun PMNs [31], and if engulfment by Mo/MØ is inhibited, may alter the conversion to an pro-reparative phenotype, and be associated with less vein wall fibrotic injury. Conversely, no difference in DDR2+ cells was found with clodronate liposomes, in contrast to the CD11b-DTR mice. We speculate that the more ‘complete’ monocyte depletion in the CD11b-DTR mice may directly affect monocyte-DDR2+ fibroblast signaling [32], with less proinflammatory mediated DDR2 fibroblast activation (such as less IL-6 in the treated mice).

The mechanisms of Mo/MØ profibrotic effects on the vein wall are not fully identified here. We did find reduced levels of IL-6, FSP-1, and VEGFα dependent on the Mo/MØ depletion strategy, but no differences in MCP-1, TGFb1 or PAI-1 expression; factors involved in other fibrotic diseases [6,11,21]. Prior work has shown reduced vein wall fibrotic injury in mice lacking IL-6 [27], but this observation did not prove a causal mechanism. We did find here that PMN depletion may affect later Mo/MØ vein wall phenotype, as lack of PMN engulfment may promote persistence of more proinflammatory Mo/MØ, with increased vein wall CCR2⁺ cells, and the associated trend with increased vein wall fibrosis. This is consistent with Mo to MØ transitions in sterile injury with engulfment of PMNs [19], possibly mediated by inhibiting proinflammatory intracellular signaling [33].

Given the lack of direct therapies for PTS, this work forms the basis for future studies to identify the kinetics of Mo/MØ influx into the vein wall, and test agents to modulate the Mo/MØ in order to positively remodel the post thrombotic vein wall.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.thromres.2026.109595.