Regenerative Medicine Therapies for Prevention of Abdominal Adhesions: A Scoping Review

Abstract

Introduction:

Globally, abdominal adhesions constitute a significant burden of morbidity and mortality. They represent the commonest complication of abdominal operations with a lifelong risk of multiple pathologies, including adhesive small bowel obstruction, female infertility, and chronic pain. Adhesions represent a problem of the entire abdomen, forming at the time of injury and progressing through multiple complex pathways. Clinically available preventative strategies are limited to barrier technologies. Significant knowledge gaps persist in the characterization and mitigation of the involved molecular pathways underlying adhesion formation. Thus, the objectives of this scoping review are to describe the known molecular pathophysiology implicated in abdominal adhesion formation and summarize novel preclinical regenerative medicine preventative strategies for potential future clinical investigation.

Methods:

A literature review was performed in accordance with the Preferred Reporting Items for Systematic Reviews Extension for Scoping Reviews. Included peer-reviewed publications were published within the last 5 y and contained in vivo preclinical experimental studies of postoperative adhesions with the assessment of underlying mechanisms of adhesion formation and successful therapy for adhesion prevention. Studies not involving regenerative medicine strategies were excluded. Data were qualitatively synthesized.

Results:

A total of 1762 articles were identified. Of these, 1001 records were excluded by the described screening criteria. Sixty-eight full-text articles were evaluated for eligibility, and 11 studies were included for review.

Conclusions:

Novel and reliable preventative strategies are urgently needed. Recent experimental data propose novel regenerative medicine targets for adhesion prevention.

Introduction

Abdominal adhesions are fibrous nonanatomic networks that develop within the abdomen after inflammation. They constitute a significant burden of morbidity, mortality, and surgical treatment globally, representing the most common complication of abdominal operation with a lifelong risk of adhesive small bowel obstruction, reoperation, female infertility, and chronic pain.¹ Studies indicate that one-fourth to one-third of laparotomy patients require rehospitalization for these adhesion-related complications within 5 y of the original insult.^2,3 The accrual of adhesiolysis-related hospital admissions accounts for annual healthcare costs of approximately $2.3 billion, requiring an average of 7.8 d of inpatient care.⁴ Barrier technologies have been used clinically to prevent the incidence of adhesion-related complications, but there is conflicting evidence to conclude if these measures correlate with improved out-comes.^5–9

Inflammation of the peritoneum, whether surgical, septic, or traumatic, is the key event initiating the molecular pathways that result in adhesion formation. In the first few hours to days after injury, acellular and cellular inflammatory mediators, along with the activated coagulation cascade, initiate the process of remesothelialization (i.e., regeneration of the peritoneum). By the end of the first week (5-8 d), the process is completed with fibroblast differentiation, collagen deposition into the extracellular matrix (ECM), and blood vessel in-growth.^10–13 Clinically available interventions for prophylaxis (i.e., Seprafilm, Baxter) mechanically limit adhesion formation between two opposing surfaces but difficulty with handling characteristics, and lack of interorgan adhesion prevention and the need for avoidance of bowel anastomoses reduce its effectiveness and utilization by surgeons.^8,14

Current efforts have shifted toward regenerative medicine therapies for the prevention of adhesion formation. Leveraging the innate healing responses of the body, regenerative medicine seeks to redirect inflammatory molecular pathways toward restorative healing and away from fibrosis.¹⁵ Two pillars, stem cells and small molecules, have great potential for clinical application in inflammatory diseases.^16,17 Pertinent to abdominal adhesion formation, a durable prevention strategy must represent an easy-to-apply treatment for the abdomen, address multiple complex pathways, and be deployed early in the disease process. Various molecular targets and interventions have been identified through a robust body of prior work, but with mixed results.¹³ Taken together, there is no consensus for adhesion prophylaxis, indicating a critical need for reproducible, clinically relevant interventions. Regenerative medicine strategies are well positioned for application to this common, morbid, and expensive problem. The following scoping review describes the known molecular pathophysiology implicated in abdominal adhesion formation and novel preclinical regenerative medicine strategies for potential future clinical investigation.

Methods

A literature search was conducted for peer-reviewed publications in accordance with the Preferred Reporting Items for Systematic Reviews Extension for Scoping Reviews guideline (PRISMA-ScR).¹⁸ A search of the MEDLINE database was performed using PubMed with the following keyword combinations (most recent search execution: July 7, 2021): postoperative adhesions, post-surgical adhesions, peritoneal adhesions, abdominal adhesions, surgical adhesions, inflammation, model, stem cells, prevention, biomarkers, pathogenesis, inhibitor, signaling pathways, small molecule, and mesothelium. Eligible articles included (1) English compositions published within the last 5 y (≥2016); (2) in vivo animal experimental models of postoperative adhesions; (3) molecular assays of implicated pathways; (4) successful mitigation of clinical adhesion formation with intervention, as quantified by the chosen adhesion clinical scoring system for each study. Nonexperimental publications, nonmolecular based therapies (i.e., barrier, recently reviewed by Tang et al.¹⁹), plant-derived therapies (recently reviewed by Soltany²⁰), interventions not involving stem cells or small molecule therapies (i.e., nonregenerative), interventions unsuccessful in adhesion reduction, malignant or congenital adhesions, and human subjects research were excluded. Abstracts were screened independently for inclusion by two of the authors (J.S. and S.P.C.) before completing full review. Discrepancies in study selection were resolved by consensus and discussion with other authors, if necessary. Screening was not biased by country of research origin. A data-charting form was developed jointly by three reviewers (J.S., S.P.C., and P.K.C.). Variables included for review (Table) were therapeutic class, drug name, animal model, survival, adhesion technique, treatment strategy, biomarker assays, mechanism of action, and author. Studies were grouped by therapeutic class (i.e., stem cells and acellular derivatives [ACDs] or small molecule inhibitors). Institutional review board approval was not required, as this review is comprised of currently published literature.

Table –

Antifibrotic strategies by therapeutic class.

Therapeutic class	Biologics/drug name	Animal model	Survival/study duration	Adhesion technique	Treatment strategy	Biomarker assays	Mechanism of action	Author
Stem cells and acellular derivatives	MSC-ACD (human bone marrow)	Mouse	1, 3, 7 d	Cecal abrasion	Single instillation at the time of surgery	↓ TNFα (peritoneal) and IFNγ (peritoneal and systemic); ↓ Mononuclear and PMN infiltration in cecum; ↓ MP (peritoneal); ↓ pericecal fibrin and myofibroblast; ↑ D-dimer (peritoneal)	Inhibition of PMN infiltration, inhibition of macrophage inflammatory products, stimulate MMP production, stimulate plasmin production, decreases fibroblast TGFβ1, collagen and ECM production, stimulates ECM degradation by MMP (1)	Rojo and Cognet 2018
	H-MSC (rat umbilical cord)	Rat	14 d	Ileal abrasion	Submucosal injection	Dose-dependent ↑ IL-10 expression in ileal tissue		Muhar et al. 2019
	MSC and M2 macrophages (rat autologous adipose)	Mouse	7 d	Cecal, peritoneal abrasion, abdominal wall suture	Preperitoneal injection of free epididymal adipose graft immediately after model creation	↑ IL-10, percentage of healed peritoneum (Keratin 8 stain)		Laukka et al. 2020
	AMD3100 (Perixafor) FK506 (Tacrolimus)	Rat	5 and 14 d	Anterior abdominal wall sutures	Subcutaneous injection, postoperative and every other day for 10 d	↑ MSC (CD133+), HGF expression, M2 macrophages within adhesion tissues	Recruitment of bone marrow–derived stem cells to sites of tissue injury (abdominal wall sutures)	Iwasaki et al. 2019
Small molecule inhibitors	Antibody (ab) to MSLN, CD47, and small molecule inhibitors	Mouse	7-13 d	Ischemic peritoneal buttons and optional organ abrasion	Anti-MSLN ab: IP injection at 7, 10, 13 d postoperatively; small molecule inhibitors: IP injection immediately postop, 4 h postop, and daily for 7 d	↓ tissue MSLN+, PDPN+ mesothelial cells after HIF1α inhibition	Inhibition of MSLN, CD47, and HIF1α pathway (2)	Tsai et al. 2018
	YC-1	Mouse	7 d	Ischemic peritoneal buttons	2 groups: Daily IP injection with 3 d pretreatment versus lavage at time of model creation	↓ tissue leukocyte/macrophage accumulation, ↓ tissue TNF-α, IL-6, iNOS (M1 markers); ↑ CCL2 (M2 marker), ↓ αSMA mRNA, ↓ peritoneal FGF, ↓ PAI-1, and VEGF	Inhibition of HIF1α pathway (3)	Strowitzki et al. 2017
	T-5224	Mouse	14 d	Cecal/abdominal wall abrasion with abdominal wall suture placement	Single application to abrasion sites at the time of surgery	↓ c-JUN, STAT, PDGFRα gene expression	AP-1 inhibitor (c-JUN pathway) (4)	Foster et al. 2020
	QLT-0267	Rat	7 d	Peritoneal abrasion	IP instillation immediately postop	↓ serum IL-6, IL-1 (culture), TNFα (culture), VEGF-A (culture)	Inhibition of integrin-linked kinase, focal adhesion kinase, and GSK-3β pathways (anticancer) (5)	Fang et al. 2018
	EW-7197	Rat	7 d	Ischemic peritoneal buttons	Oral gavage for 7 d postop + vehicle only for 21 d	↓ collagen, TGFβ, SMAD 2/3	TGFβ1-receptor kinase inhibitor (6)	Tsauo et al. 2018
	Pirfenidone	Rat	14 d	Cecal abrasion	Oral gavage for 14 d postop	↓ TIMP-1, TNFα, TGFβ1; ↑ MMP-9	Anti-inflammatory, antifibrotic (7)	Bayhan et al. 2016
	Pentoxifylline (PTX)	Mouse	7 d	Cecal abrasion	IP injection 2 d preop and 7 d postop	↓ Collagen, ki67+/CD31+ cells, F4/80, FSC1, αSMA cells, ↑ tPA	PDE inhibitor (8)	Yang et al. 2018

CCL2 = Chemokine (C-C Motif) ligand 2; FGF = fibroblast growth factor; GSK-3β = glycogen synthase kinase; iNOS = inducible nitric oxide synthase; MP = macrophage; PO = per oral; PDE = phosphodiesterase.

Results

Overall description of included studies

In compliance with the PRISMA-ScR guidelines, we identified 1762 articles by database interrogation. Of these, 693 records were excluded for duplication, and 1001 records were excluded by the described screening criteria. Sixty-eight full-text articles were evaluated for eligibility, and 11 studies were included for a review (Fig. 1).

Fig. 1 –

Electronic database search results in accordance with PRISMA-ScR guidelines.

Study characteristics and predictors

The selected 11 studies introduced potential regenerative therapies for adhesion prophylaxis. These were categorized into two groupings by therapeutic class: stem cells and ACDs or small molecule inhibitors (Table). There were four articles that evaluated stem cell therapies and seven that evaluated small molecule inhibitors. Cell-based therapies used stem cells from various sources, including human bone marrow, rat umbilical cord, rat adipose tissue, and recruitment of endogenous rat bone marrow cells. Regarding small molecule therapies, sources included experimental molecules as well as clinical treatments for cancer, fibrosing, and inflammatory diseases. All experiments used rat or mouse models. To assess adhesion development, animals were survived for a maximum of 7-14 d (mode, 7 d). To induce adhesions, the authors used previously described abrasion techniques, most commonly to the cecum and abdominal wall, but also to ileum, and uterine horn.^19,21 Ischemic peritoneal buttons, in which a suture ligature was tied around a portion of parietal peritoneum to induce local hypoxia and tissue injury, was used by multiple studies (n = 5). Treatment strategies varied but most commonly consisted of intraperitoneal (IP) therapies at the time of operation and continuing postoperatively via IP injection for varying periods. Other strategies of therapy administration included submucosal, subcutaneous, or enteric delivery via oral gavage.

There were multiple measures (i.e., scoring systems) of macroscopic adhesion severity used in the included studies. Consistent with the eligibility criteria, all authors reported successful macroscopic (±microscopic) reduction in adhesions, quantified by these various scoring systems described in detail elsewhere.^22–26 Each scoring system contained differing criteria for adhesion grading and degrees of subjectivity. None encompassed a comprehensive assessment of the entire abdomen. Measured biomarkers of adhesion formation were commonly proinflammatory cytokines (tumor necrosis factor [TNF]-α, interferon-gamma [IFN-γ], and interleukin [IL]-6), profibrotic growth factors (tumor growth factor [TGF]-β1, fibroblast growth factor, platelet-derived growth factor [PDGF], and vascular endothelial growth factor [VEGF]) and associated proteins, including alpha smooth muscle actin (α-SMA), collagens, tissue inhibitor of metalloproteinase-1 (TIMP-1), matrix metalloproteinases (MMP), plasminogen activator inhibitor-1 (PAI-1), and small mothers against decapentaplegic (SMAD) family of signal transduction proteins. Anti-inflammatory agents, such as IL-10 and hepatocyte growth factor (HGF), were also measured. Neutrophils, lymphocytes, macrophages, and their associated phenotypes were assessed. Profibrinolytic proteins, such as tissue plasminogen activators (tPA) along with fibrinolysis (fibrin and fibrin split products), were also measured. Experiments with small molecules included gene expression assays for transcription factors implicated in fibrosis (hypoxia-inducible factor 1 alpha [HIF1α] and c-JUN). Specific biomarker changes associated with adhesion reduction are reported in Table.

Discussion

Mechanisms of peritoneal fibrosis

Immune response

Immediately after abdominal operation, acellular and cellular inflammatory mediators of the immune response initiate tissue repair of the damaged peritoneum (i.e., remesothelialization; Fig. 2). This course is thought to complete approximately 1 wk after injury.¹⁰ These processes advance under tissue hypoxic conditions, sensed by the HIF1α pathway, under which antiapoptosis, mesothelial-to-mesenchymal transition (MMT, also known as epithelial-to-mesenchymal transition), angiogenesis, and fibrosis occur within an ECM substrate.^{10,12,13,27,28} Notably, MMT is defined by the loss of cell–cell adhesions with downregulation of E-cadherin, acquisition of mesenchymal cell (spiculated) morphology, and increased contractility of actin fibers.²⁹

Fig. 2 –

Mechanisms of abdominal adhesion formation. Numbers in gray boxes correspond to parenthetical numbers under the Mechanism of action column in Table. PAF = platelet-activating factor; ROS = reactive oxygen species; uPA = urokinase plasminogen activator; FDP = fibrin degradation product; ROCK = Rho-associated protein kinase.

These processes are communicated by multiple signals. TNF-α and IL-1, specifically IL-1β, and IL-6 are proinflammatory mediators, produced by macrophages and mast cells during the acute inflammatory phase. Their overexpression has been associated with end-organ fibrosis for a variety of models in dose-dependent fashion.^19,28 TGF-β1, also a product of macrophage activation, displays similar profibrotic properties and is also a central controller of mesothelial cell transition to myofibroblasts.²⁸ Interestingly, TGF-β1 has pleiotropic effects (both pro- and anti-inflammatory) based on its cell of origin and impacts multiple biologic pathways, making its utility as a therapeutic target quite challenging.²⁸ IFN-γ is a product of activated T helper cells, and although considered proinflammatory by virtue of activating macrophages to a classic (M1, inflammatory) phenotype, it may also inhibit fibrosis by antagonizing the activity of TGF- β1 and fibroblast proliferation.²⁸

As in other forms of fibrosis, myofibroblasts are an essential cell type in the formation of abdominal adhesions. Until recently, the cell of origin for peritoneal myofibroblasts remained in question. Tsai et al. used multiple lineage tracing strategies to demonstrate that the peritoneal mesothelium responds to injury by undergoing gene expression changes (~8000 genes at 24 h after injury). These changes allow it to take on a myofibroblastic phenotype with loss of E-cadherin and production of α-SMA, myosin, collagen, and ECM deposition.³⁰ Although other cell types, such as circulating fibrocytes and submesothelial fibroblasts, could play some role in adhesion generation, the majority contribution arises from mesothelial cells following stress activation.^28,30–33 Of great interest are the findings recently demonstrated by Fischer et al., who not only reconfirm the mesothelium to be the cell of origin for abdominal adhesions but also show that the “pathologic cell phenotype” (i.e., myofibroblast) may be induced in healthy surfaces by direct contact with injured cells.³³

Indeed, previous work demonstrates the MMT to be a foundational step in the genesis of abdominal adhesions.^34,35 Both Tsai and Fischer recognized that major transcriptional changes reflective of this phenotype take place within the first 6-8 h after injury.^30,33 As demonstrated in Figure 2, implicated intracellular transduction pathways in MMT include (1) ERK (extracellular signal-regulated kinase, also known as mitogen-activated protein kinases [MAPK]) and SMAD2/3, activated by TGF-β1; (2) nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) activated by IL-1; (3) Janus kinases-signal transducer activator of transcription (JAK-STAT) activated by IL-6.³⁴ Of note, there is the potential for significant overlap between intracellular pathways, as exemplified by PDGF, which uses ROCK (Rho-associated protein kinase), NF-κB, JAK-STAT, and ERK.³⁶ Genes implicated in these profibrotic pathways include Snail (zinc-finger protein SNAI1), Twist (twist-related protein), α-SMA, PAI-1, collagen family, fibronectin, MMP2 and MMP9 with inhibition of E-cadherin, and Wilms tumor protein 1.³⁴

Coagulation response

In tandem with the immune response is the immediate and persistent activation of the coagulation response, resulting in an uninterrupted ECM. Similar to the immune response, the coagulation response begins with factor recruitment by degranulation of mast cells. Histamine and platelet-activating factors increase vascular permeability, leading to localization of fibrinogen, fibronectin, and platelets. Comparable to macrophages, TGF-ß1, IL-6, and PDGF are secreted by platelets as chemoattractants to inflammatory cells and activators of MMT.^13,28,37 A polymerized gel matrix simultaneously results from activation of the coagulation cascade, with the conversion of fibrinogen to fibrin by thrombin. Under nonpathologic conditions, counterbalance is facilitated by innate fibrinolysis from tPA, secreted by mesothelial cells and macrophages, producing fibrin degradation products. However, this mechanism is inhibited under tissue hypoxic conditions by fibroblast production of TIMP-1 and PAI-1 and 2, shifting the balance toward antifibrinolysis.¹³ This ratio has demonstrated clinical importance, as high concentrations of endogenous PAI-1 and low tPA activity are observed in tissue biopsies of patients with extensive peritoneal adhesions.³⁸ In the later phases of fibrogenesis (i.e., fibrosis), the myofibroblast phenotype predominates with collagen deposition and maturation of the persistent fibrin gel matrix. VEGF, initially released by mast cells during the acute inflammatory phase, promotes angiogenesis along with TGF-ß1 during fibrosis.¹³

Multiple modalities have been previously used to interrupt these complex pathways in the development of abdominal adhesions. Prophylaxis by surgical technique includes avoidance of unnecessary dissection, prevention of gastrointestinal spillage and laparoscopic approach, all of which have been shown to decrease fibrinogenesis.^5,39 A variety of antiadhesive barriers have been developed to introduce an absorbable and inert “spacer” between two injured surfaces of peritoneum, thus interrupting the formation of adhesions.¹⁹ Although collectively these technologies have taken on multiple forms (i.e., solid barrier films, gels, and liquids) and follow an intuitive rationale, the complicated mechanisms of adhesion generation overwhelm this concept as a single strategy, resulting in partial and/or inconsistent reductions in adhesion formation. Furthermore, difficulty with handling characteristics and the need for avoidance of bowel anastomoses due to morbidity makes for further challenges with routine use of barrier technologies.^8,40–42 Therefore, the purpose of this review is to summarize current promising regenerative-based interventions in preclinical studies.

Regenerative Medicine Therapies

The field of regenerative medicine integrates the technologies of cell transplantation, materials science, and bioengineering to achieve disease reversal and tissue restoration.⁴³ Therefore, regenerative therapies are built on the concepts of cellular repair in acute inflammation and rejuvenation in chronically injured organs. As a result, stem cells, fundamental to the regenerative medicine landscape, along with their anti-inflammatory products, function as promoters of normative healing and potential precursors to cell types of the target organs. While early work in regenerative medicine considered stem cells primarily to be direct replacements of various tissues, the evolution of understanding has directed focus to their paracrine function as significantly responsible for their healing potential.⁴⁴ Small molecules, with their capacity for redirecting the many complex pathways of intracellular signaling, are well positioned to explore these relationships while also modifying the direction of healing over time.¹⁶

Taken together, stem cell and small molecule–based therapies are the current focus of regenerative-based molecular interventions for abdominal adhesions. Either modality represents an opportunity, whereby the inflammatory framework may be redirected toward the restoration of a nonpathologic anatomy. Although the objective of abdominal adhesion reduction may be shared between regenerative and traditional molecular medicine, regenerative medicine extends beyond the goal of limiting fibrosis to guide multiple pathways toward normal healing.

Stem cells

To date, stem cell therapeutic applications are best established in lymphoid and myeloid malignancies with hematopoietic bone marrow transplantation and in hematologic disorders with umbilical cord blood.^15,45 For non-hematomalignant disease processes, mesenchymal stem cells (MSCs), isolated from adult sources (i.e., bone marrow and/or adipose), produce an anti-inflammatory secretome that aids in repair of damaged tissues due to ischemia, inflammation, and autoimmunity.¹⁵ As recently described by Almeida-Porada et al., MSCs possess several features that favor their role in regenerative medicine strategies, namely, ease of tissue procurement, ease of expansion with functional fidelity, host immune neutrality, and a honing ability to damaged tissues.⁴⁶ Beyond bone marrow and adipose, a ready source of MSC is presented in the form of gestational tissues, such as the placenta, placental membranes, umbilical structures, and amniotic fluid.¹⁷ These tissues are rich in multipotent cellularity, possessing advantages over alternative sources because of quicker expansion, minimal immunogenicity, and maximal anti-inflammatory effect.¹⁷ Recent evidence for this observation is provided by Khoury et al., who demonstrate that co-culture of bone marrow stromal cells derived from the amniotic fluid stem cells and human placental stem cells with lipopolysaccharide-stimulated leukocytes resulted in decreased gene expression of IL-1ß, IFN-γ, TNF-α, neutrophil elastase and NF-κB.⁴⁷ Of these groups, human placental stem cell demonstrated the greatest inhibitory effect on NF-κB and neutrophil elastase gene expression.

Rojo and Cognet have recently shown that administration of IP MSC ACDs mitigates abdominal adhesion formation.¹¹ ACDs consist of proteins, messenger RNA, and microRNA, all of which may be both soluble and contained within microvesicles. In a murine cecal erosion model, ACDs cultured from human bone marrow MSC were instilled once at the time of injury creation. Animals were survived for 7 d and demonstrated reductions in the macroscopic severity of adhesions at necropsy.¹¹ This finding was associated with a significant decrease in peritoneal myofibroblasts, pericecal fibrin, peritoneal and systemic proinflammatory cytokines and infiltration of monocytes and polymorphonuclear (PMN) cells into the cecal mesothelium. Furthermore, they observed significantly increased levels of peritoneal fibrin split products (D-dimer), consistent with breakdown of the fibrin gel matrix. The authors described multiple potential bioactives within the MSC-ACD (Fig. 2), including (1) IL-10, an anti-inflammatory cytokine responsible for decreasing neutrophil infiltration, decreasing TGF-ß1 production by macrophages, and promoting MMP production by fibroblasts; (2) tPA and urokinase-type plasminogen activator, which facilitate the conversion of plasminogen to plasmin and the digestion of the fibrin gel matrix; (3) HGF, which reduces the production of TGF-ß1 and collagen types I and III by fibroblasts, possibly via inhibition of SMAD2/3, while promoting synthesis of MMP (subtypes 1, 3, 13); (4) adrenomedullin, an inhibitor of fibroblast ECM production; and (5) MMPs, enzymes affecting degradation of the ECM.^11,34

Additional experiments by Muhar et al. used rat umbilical cord MSC that were cultured in hypoxic conditions (H-MSC, 5%O₂).⁴⁸ Hypoxia, when applied to MSC, increases gene expression of multiple molecules involved in stem cell metabolism, proliferation, and survival, including VEGF, HGF, adrenomedullin, MMP9, and prostaglandin E synthetase (PGE).^49,50 Of note, TNF-α, IFN-γ, IL-12, and IL-6 were all observed to decrease expression in co-cultured macrophages, possibly due to elevated PGE2 production with induction of an anti-inflammatory (M2) macrophage phenotype and activation of regulatory T cells (PGE2-M2-Treg axis).⁵⁰ This observation is in contrast to the upregulation of proinflammatory cytokines by monocytes when cultured under similar conditions.^49,50 H-MSC at two concentrations were then injected submucosally following model creation, resulting in decreased adhesion formation and dose-dependent increases in IL-10 expression.⁴⁸ In related experiments, Iwasaki et al. stimulated recruitment of endogenous bone marrow–derived MSC, identified by CD133, to injured peritoneal sites with subcutaneous injection of AMD3100 (Plerixafor) and FK-506 (Tacrolimus) every other day for 10 d.⁵¹ The investigators concluded that reductions in clinical adhesion scores were associated with injury localization of MSC, induction of an anti-inflammatory (M2) phenotype, and production of the relevant secretome (i.e., IL-4, IL-10, and HGF) among macrophages.⁵¹ In agreement with Rojo and Cognet, the authors endorse the role of HGF in the prevention of adhesions, specifically through inhibition of IFN-γ and PAI-1 with induction of mesothelial cell proliferation.^52–54

In a separate approach, Laukka et al. found that autologous fat grafting, used for hypertrophic scar treatment because of immune-modulatory effects, can also prevent peritoneal adhesions.⁵⁵ Immediately after model induction with abrasion to the peritoneum and cecum, investigators performed preperitoneal injection of a free epididymal fat graft. They concluded that fat grafting can inhibit adhesion formation via mesothelial regeneration and that upregulation of IL-10 is critical to this process.⁵⁵ These findings are in agreement with prior work by Holschneider et al. establishing the role of IL-10 in adhesion prevention.⁵⁶ IL-10, produced by anti-inflammatory (M2) macrophages, regulatory T and Th2 lymphocytes, has also been used to induce macrophage polarization from an M1 to M2 phenotype with subsequent reductions in NF-κB pathway activation and TNF-α production.^57,58 As IL-10 is a product of MSC, the mechanism of M2 phenotype induction with subsequent tissue healing, as proposed by Iwasaki et al., is further supported in this set of experiments.⁵¹

In contrast to these studies, Karaca et al. showed that adipose-derived MSCs, applied topically to the cecum following sponge abrasion, decreased circulating concentrations of PMN, adhesion molecules (E- and P-selectin) and inflammatory cytokines (TNF-α and IL-1) without clinical reduction in adhesions.⁵⁹ The investigators report a small sample size to be a major limitation in their lack of statistical significance. Similarly, Uysal et al. present negative clinical and microscopic data, revealing that allogeneic adipose-derived stem cells introduced to the abdomens of rats following laparotomy with cecal abrasion resulted in worse adhesion scores and tissue histopathologic markers (i.e., collagen deposition, E-cadherin).⁶⁰ Of note, the dose of stem cells administered in this study was nearly 10-fold less than in other studies discussed.

Future directions in cell-based therapies

Preclinical models for abdominal adhesion reduction using regenerative medicine therapies have been in existence for over 20 y, yet many questions still remain.⁶¹ Cell-based therapies act on multiple therapeutic targets, encouraging an anti-inflammatory (M2) macrophage phenotype,^51,62 decreasing Th1 lymphocyte proliferation⁴⁷ and discouraging the MMT.⁶³ Nonetheless, practical challenges associated with clinical translation of stem cell-based therapies remain in the realms of cell line standardization (i.e., optimal stem cell type and culture conditions), dosing protocols, timing of delivery, and route of delivery. These variables in the current literature vacillate widely, making comparison between studies difficult. Although applications of MSCs appear to be relatively safe, therapies are also not without risks, including the potential for tumorigenicity, thrombo-embolic events, and relegation within end-organ beds, if administered systemically.^64–67 While additional concerns may arise due to wound and anastomotic healing, the available data evaluating stem cell therapies in this regard suggest a potential benefit.^68–71 Although the therapeutic advantages of stem cells are accomplished through multiple avenues, future evaluations will undoubtedly expand on the details of these mechanisms, the role of other stem cell ACD (i.e., exosomes) and cross-talk between cell lines (i.e., macrophages, lymphocytes, and neutrophils) involved in the regulation of healing.^72–74

Small molecules

Small molecules are any variety of organic compound with low molecular weight. Common examples are aspirin, diphenhydramine, penicillin, and atorvastatin. Advantages of small molecules include a simple chemical structure, predictable chemistry, and straightforward dosing protocols.⁷⁵ Furthermore, they are often easier to manufacture and move through the regulatory process, an advantage over biologic class medications.⁷⁵ Small molecules have found application in a variety of end-organ fibrosing disease processes, including skin, lung, liver, and kidney.⁷⁶ As they pertain to regenerative medicine, small molecules are well positioned to promote the production of target cells in vitro and to stimulate the proliferation of endogenous stem cells in vivo.¹⁶ As in other fibrosing disease processes, small molecules are able to be directed toward specific targets involved in adhesion formation, allowing them to redirect signaling within critical pathways. Of these, several recent reports have highlighted the roles of HIF1α and c-JUN transcription factors in pathways essential to the generation of abdominal adhesions.^30,37,77,78

Hypoxia-inducible factor 1-alpha

The HIF1α, a transcription factor subunit coded by the HIF1α gene, is fundamental to oxygen sensation and adaptation in mammalian life, a discovery that resulted in the 2019 Nobel Prize for Physiology or Medicine.⁷⁹ The subunits of HIF (α and ß) are both expressed under circumstances of tissue normoxia and hypoxia. However, under tissue hypoxia, the inhibitor of HIF1α loses its ability to recognize the HIF1α substrate, allowing for dimerization with its beta subunit and migration into the nucleus with transcription induction of the HRE (hypoxia response element) gene. HRE is responsible for the regulation of multiple oxygen-responsive genes, including those of cellular metabolism, erythropoiesis, angiogenesis (i.e., VEGF), and MMT (i.e., Snail1, Twist2).^78,79

Tsai et al. recently demonstrated the role of HIF1α in the genesis of abdominal adhesions.³⁰ First, via cell lineage tracing analysis, the investigators showed that mesothelin (MSLN), a fetal mesothelial cell surface marker, is specifically upregulated (80-fold) among noncirculating mesothelial subpopulations (i.e., podoplanin positive, PDPN+) participating in adhesion formation. Treatment with antibodies against MSLN and CD47, a “no-kill” surface antigen, induced an immune-mediated elimination of these cell types, significantly decreasing adhesion generation in a murine model. Among the roughly 8000 genes that underwent differential expression following model induction, the HIF1α pathway was observed to increase expression to peak 6 h postoperatively in MSLN+/PDPN+ mesothelial cells in vitro (Fig. 3). Subsequent inhibition of this pathway with small molecules, directed to various aspects of HIF1α, significantly reduced peritoneal adhesion formation in the mouse at 7 d, with many animals demonstrating no adhesions.³⁰ Furthermore, small molecule inhibition of HIF1α resulted in M2 macrophage polarization, decreased MMT, promotion of fibrinolysis, and inhibition of angiogenesis, as described by Strowitski et al.⁷⁸ Taken together, this process of “mesothelial cell reprogramming” to a fibroproliferative MSLN+ subtype appears to be mediated in part by the HIF1α signaling pathway. Additional critical findings in these sets of experiments were (1) animals that received inhibitor treatment appeared healthy and demonstrated normal wound healing; (2) samples of human intra-abdominal adhesions revealed similar staining patterns, revealing MSLN and CD47 positivity; (3) mesothelial cells cultured under hypoxia in the absence of macrophages did not demonstrate fibrotic foci, suggesting the essential cross-talk between mesothelial cells and macrophages in the generation of adhesions.^30,78

Fig. 3 –

Timing and concentrations of cell types within peritoneum with respect to HIF1α and c-JUN activation postoperatively. Modified from Tang et al. Peritoneal adhesions: Occurrence, prevention and experimental models. Acta Biomater. 2020; 116:84-104. https://doi.org/10.1016/j.actbio.2020.08.036.

c-JUN

The protein c-JUN, a component of the transcription factor complex activator protein-1 (AP-1), has been centrally implicated in a variety of human fibrosing diseases, including peritoneal adhesions.⁷⁷ c-JUN is impacted by the transduction pathways of VEGF, fibroblast growth factor receptor, platelet-derived growth factor receptor (PDGFR), and transforming growth factor beta receptor.⁷⁷ As hypothesized by Foster et al., IL-6 and PDGF secreted by platelets localizing to inflamed tissues stimulate the activation of c-JUN with subsequent amplification of profibrotic pathways (Fig. 2).³⁷ The authors showed that c-JUN is activated by JAK-STAT and MAPK (i.e., ERK) after PDGFRα stimulation, resulting in gene expression of profibrotic FSP1, MCP1, IL6, STAT, ASMA, and MMT genes.³⁷ Building on the work of Tsai et al., c-JUN expression was correlated with MSLN expression, indicating a potential relationship between c-JUN and HIF1α (Fig. 3).³⁷ Indeed, previous investigations suggest that HIF1α and c-JUN cooperate in signal transduction pathways, with c-JUN protecting HIF1α from degradation.⁸⁰ This observation is further evidenced by recent experimental studies in which cardiac fibrosis was mitigated in association with reduced gene expression of left ventricular c-JUN, HIF1α, and VEGF.⁸¹

Interest in the HIF1α pathway, as it pertains to abdominal adhesion formation, has been ongoing since the early 2000s.⁸² That said, the genetic and epigenetic targets modified by c-JUN and HIF1α remain to be fully described. As an example, Wilms tumor protein 1, a marker of MMT, is described as a potential target by Tsai et al., but this was not demonstrated in the subsequent work by Foster et al.^30,37 In addition, there appears to be a provocative relationship between HIF1α and c-JUN that requires further elaboration. Both factors play a contributory role to the MMT, and the available data suggest at least a partial overlap in their timelines (Fig. 3).⁸³ Furthermore, there appears to be multiple intracellular signaling pathways by which c-JUN (i.e., AP1) activates the MMT, including JAK-STAT, MAPK (i.e., ERK1/2) and ROCK (Rho kinase), and NF-κB.^84,85 Importantly, two of the reviewed studies showed no negative effects of small molecule HIF inhibitors on wound healing, suggesting nonoverlap between the two pathways and a promising target for future clinical application.⁷⁸

Other small molecule targets

Several other studies have used small molecules to inhibit abdominal adhesion formation and prevent MMT. Fang et al., in evaluating the possible proinflammatory effect of fibrin on the peritoneum, blocked the effects of fibrin-induced cytokine production on mesothelial cells with QLT-0267, an integrin-linked kinase inhibitor.⁸⁶ The investigators showed that integrin-linked kinase inhibition decreased adhesions coincident with decreased cytokine (IL-6 and IL-1β) production through the GSK-3β and NF-κB pathways. Of note, fibrin was previously shown by this group to be an inducer of MMT.⁸⁷ Tsauo et al. used EW-7l97, an oral TGF-ß1 receptor kinase inhibitor currently being tested as an anticancer therapy, in a rat ischemic peritoneal button model via oral gavage for 7 d postoperatively.⁸⁸ They reported decreased SMAD2/3 and αSMA and increased E-cadherin within peritoneal buttons, also favoring the prevention of MMT.⁸⁸ Bayhan et al. used pirfenidone, a small molecule cytokine inhibitor treatment for idiopathic pulmonary fibrosis, in a rat cecal erosion model as an oral gavage for 14 d postoperatively.⁸⁹ Although the mechanisms of pirfenidone remain incompletely characterized, investigators observed decreased TIMP-1, TNF-α, and TGF-β1 with elevations in MMP9, suggesting its role in preventing the myofibroblastic phenotype.^89,90 Furthermore, additional data propose a role for pirfenidone in the treatment of insulin resistance via M2 macrophage polarity.⁹¹ Finally, Yang et al. used pentoxifylline, a non-specific phosphodiesterase inhibitor, in a murine cecal erosion model through IP injection 2 d preoperatively and continued 7 d postoperatively.²⁹ They reported marked decrease in adhesions, supported by decreased collagen deposition and adhesion thickening. Other associated findings included increased tPA and decreased number of ki67+/CD31+ cells, F4/80+, FSP-1+, and αSMA+ cells at the adhesion site, leading the authors to conclude a decrease in the myofibroblastic phenotype.²⁹

Conclusions

Abdominal adhesions generate a spectrum of disease with significant morbidity and healthcare costs. Mechanistically sound interventions for effective and reproducible prophylaxis of the entire abdomen are a surgical imperative. Current clinically available interventions are either cumbersome, expensive, and/or marginally effective. Novel approaches are urgently needed. The most recent data suggest the mesothelium to be the cell of origin for adhesions.³³ However, subsequent disease progression represents a complex interplay between macrophages, lymphocytes, and mesothelial cells, resulting in the MMT and myofibroblastic phenotype.

While the available data are promising, our scoping review should be approached in the context of some limitations. Although all investigations demonstrated successful adhesion reduction in animal models, questions remain about translation into human adhesion formation and prevention. Nonetheless, several groups present supporting data from human tissues.^30,37 The heterogeneity of the animal models, study designs, assays, treatment protocols, and scoring systems did not allow for a meta-analysis of these data. Thus, qualitative approaches for results evaluation were undertaken, represented in Table.

The shortcomings of the clinically available interventions are those of concept. With immediate and redundant amplification at each level of signaling, barrier technologies, as a single measure, are inadequate for a satisfactory outcome. Furthermore, as activated mesothelial cells may induce a pathologic phenotype in healthy mesothelium, adhesions represent a problem of the entire abdomen and require a comprehensive peritoneal solution. In lieu of a single target, a guiding principle may be to facilitate an environment within the abdomen whereby dysregulation is redirected toward regeneration of normal anatomy. As such, the regenerative medicine framework, in the forms of stem cell therapies, material science, and small molecule approaches, is well positioned to facilitate appropriate intercellular communication for translation into clinical healing.