Limited Treatment Options for Diabetic Wounds: Barriers to Clinical Translation Despite Therapeutic Success in Murine Models

Abstract

Significance: Millions of people worldwide suffer from diabetes mellitus and its complications, including chronic diabetic wounds. To date, there are few widely successful clinical therapies specific to diabetic wounds beyond general wound care, despite the vast number of scientific discoveries in the pathogenesis of defective healing in diabetes.

Recent Advances: In recent years, murine animal models of diabetes have enabled the investigation of many possible therapeutics for diabetic wound care. These include specific cell types, growth factors, cytokines, peptides, small molecules, plant extracts, microRNAs, extracellular vesicles, novel wound dressings, mechanical interventions, bioengineered materials, and more.

Critical Issues: Despite many research discoveries, few have been translated from their success in murine models to clinical use in humans. This massive gap between bench discovery and bedside application begs the simple and critical question: what is still missing? The complexity and multiplicity of the diabetic wound makes it an immensely challenging therapeutic target, and this lopsided progress highlights the need for new methods to overcome the bench-to-bedside barrier. How can laboratory discoveries in animal models be effectively translated to novel clinical therapies for human patients?

Future Directions: As research continues to decipher deficient healing in diabetes, new approaches and considerations are required to ensure that these discoveries can become translational, clinically usable therapies. Clinical progress requires the development of new, more accurate models of the human disease state, multifaceted investigations that address multiple critical components in wound repair, and more innovative research strategies that harness both the existing knowledge and the potential of new advances across disciplines.

Luisa A. DiPietro, DDS, PhD

Lin Chen, MD, PhD

Scope and Significance

This text reviews the advances in diabetic wound healing research and discusses the obstacles preventing them from successfully progressing to the clinic. The discussion focuses on research using biologically active substances to improve wound healing in murine models of diabetes mellitus (DM).

Approaches relying exclusively on wound care tools, such as inactive dressings, LED lights, bandages, nanoparticles, and gels; as well as preexisting Food and Drug Administration (FDA)-approved ther apies for other conditions; genetic modifications; and physical manipulation of wounds, such as splints, stitches, modified pressure environments, or external structures, are outside the scope of this review and will not be discussed in depth.

Translational Relevance

For the millions of patients affected by DM, impaired wound healing is a life-threatening complication that generates massive social and economic burdens.¹ To date, few successful therapies for diabetic wounds have reached the clinic² despite a vast breadth of research demonstrating improved healing in murine models of the disease. In this work, we summarize these major advances and discuss the remaining challenges that must be addressed to translate these discoveries into successful therapeutics for diabetic ulcers.

Clinical Relevance

Patients with DM face a lifetime of continuous disease management, which is further complicated by sequelae such as the diabetic foot ulcer (DFU). Currently, clinical management of diabetic wounds relies on frequent foot examination, regular care, early detection, wound debridement, appropriate dressings, pressure off-loading, treatment of infection, and, sometimes, skin substitutes.² Despite these options, diabetic wounds remain a clinical need for which there are few specifically targeted pharmacological or bioactive drugs. As research in murine diabetic models has progressed, promising potential therapeutics have often failed to reach the clinic, leaving high-risk patients with limited options.

Background/Overview

DM is a global health care concern that affects nearly half a billion people and causes over 1 million deaths annually.³ Many can be attributed to the serious complications of DM, which include heart disease, stroke, neuropathy, nephropathy, peripheral artery disease, retinopathy, and poor wound healing.^4,5 Of these complications, chronic lower extremity wounds (diabetic foot ulcers, or DFUs) are a leading contributor to the morbidity and mortality of DM, leading to over 100,000 annual amputations in the United States alone and generating a substantial social, economic, and health care burden.^1,6

The pathogenesis of deficient wound healing in DM is complex and multifactorial. Contributing factors to the inability of diabetic skin to heal include poor perfusion and angiogenesis; hypoxia; increased inflammation and tissue destruction; increased levels of matrix metalloproteinases (MMPs), free radicals, and advanced glycation end products; and dysregulated cell function.^2,5,7,8 Together, these factors predispose the skin to injury and decrease its baseline ability to heal, resulting in chronic wounds that become limb- and life-threatening (Fig. 1).

Figure 1.

Examples of diabetic wounds. Representative images of nonhealing neuropathic, ischemic, or neuroischemic diabetic ulcers affecting the lower extremities. All images generously provided by Dr. Stephanie Wu, DPM, MSc, FACFAS; Dean of Dr. William M. Scholl College of Podiatric Medicine; Professor, Department of Podiatric Surgery and Applied Biomechanics; Professor, Center for Stem Cell and Regenerative Medicine; and Director, Center for Lower Extremity Ambulatory Research (CLEAR) for Dr. William M. Scholl College of Podiatric Medicine at Rosalind Franklin University of Medicine and Science. Color images are available online.

The wound healing process consists of four specifically regulated phases: hemostasis, inflammation, proliferation, and remodeling (Fig. 2).^9,10 During hemostasis, blood vessels constrict in response to injury, and the activation and aggregation of platelets forms a blood clot known as the platelet plug, which seals off the injured area to prevent continued bleeding. The inflammatory phase then begins in response to the waves of cytokines and chemokines released during hemostasis. This phase consists of the recruitment of immune cells over the next few hours to days.

Figure 2.

Stages of wound healing. A schematic (not to scale) depicting the four phases of wound healing: hemostasis, inflammation, proliferation, and remodeling. Artwork and renderings in this and subsequent figures are courtesy of Smart Servier Medical Art, freely available at https://smart.servier.com Color images are available online.

First, neutrophils arrive to clear debris and microbes from the injured area. Although they are effective at preventing infection, the enzymes and byproducts generated by neutrophils can often cause additional tissue destruction. Next, monocytes migrate to the wound and differentiate into macrophages, which remove apoptotic cells and debris, and, later, promote fibroblast and keratinocyte proliferation and angiogenesis. As the inflammatory phase resolves over the next few days, the proliferative phase begins to replace and repair the injured tissue. This phase overlaps with the resolution of inflammation and involves the formation and deposition of new epithelial layers, granulation tissue, extracellular matrix, and many new capillaries.

Finally, the healing wound enters the remodeling phase, a long-term and ongoing process responsible for vascular pruning, remodeling of the extracellular matrix, and an overall biological “revising and editing” of the tissue to resemble normal tissue as closely as possible.^9,10

In diabetic skin, the wound healing paradigm is disrupted: whole tissue dysfunction contributes to increased injury and impaired healing; the phases of wound healing do not occur or resolve appropriately; individual cells and functions are abnormal; and the increased susceptibility of diabetic patients to external stressors, such as infection, further interferes with wound resolution.⁷ While these factors are prominent players in the diabetic wound healing problem, they are by no means the exhaustive list.

As scientists understand more about DM itself and about tissue repair and regeneration, their toolkit and the complexity of their questions continue to expand. Scientists have developed animal models to study diabetic wounds, begun deciphering the innumerable factors at play, and described hundreds of targets that improve healing in animal models of DM.

Despite this breadth of progress, however, current research advancements have produced few practical and usable clinical therapeutics for patients with DFUs. This lack of treatment options can be attributed, in part, to the animal model systems of diabetic wound healing, which have yet to sufficiently replicate the human condition—a concern that further underscores this disparity. Still other challenges arise in the complexity of the patient population, which cannot be fully represented in a clinical trial cohort, and the narrow scope of many research studies, which reveal valuable insights into the physiology of the diabetic wound but often do so only a single factor at a time.

The imbalanced ratio between preclinical research and clinically successful therapeutics leaves hundreds of thousands of patients with few options beyond standard wound care. For those with DM, the tenets of ulcer prevention and treatment are good glycemic control, daily foot care, regular examination, early detection of injuries, infection control, and appropriate wound care and dressings. Appropriate care can include wound debridement, systemic antibiotics, revascularization procedures, and pressure off-loading (Fig. 3).^2,11

Figure 3.

Current standards of diabetic foot care. A summary of the most important components of diabetic foot care and the treatment and prevention of diabetic foot ulcers. These are good glycemic control, regular foot care and examination, early detection, and appropriate wound care and dressings. There are currently no gold standard pharmaceutical therapies that are specific for and widely usable to treat diabetic wounds, although there are a few FDA-approved options available. FDA, Food and Drug Administration. Color images are available online.

Beyond these standards, wound care clinicians have only a single FDA-approved drug specific to DFUs in their arsenal: Regranex (becaplermin), a topically applied human recombinant platelet-derived growth factor-BB (PDGF-BB) gel^12,13 approved for neuropathic diabetic ulcers, which are well-perfused and do not encompass the full complexity of most DFUs. Regranex showed significant efficacy in clinical trials with tightly controlled wound criteria and treatment conditions^14,15; however, its clinical use has yet to match up to its suggested potential.¹⁶

In addition to Regranex, three biologic medical devices are FDA-approved for the treatment of DFUs. These include Dermagraft^17–19 and Apligraf (also known as Graftskin),^20–24 both of which are cell-based human skin equivalents or substitutes (HSEs), and Omnigraft, an animal-derived matrix originally used for burns.^25–27 While these are the only bioactive medical devices that are approved specifically for use in DFUs, there are many general wound care products and therapies, including other FDA-approved devices; many different types of dressings, scaffolds, bandages, sprays, and closure devices; and drugs with off-label usage; these are all used to varying degrees of success in the management of DFUs.^28–32

Throughout this review, various treatment types will be discussed, including pharmacologically active drugs, such as Regranex; medical devices, such as the HSEs; and natural or inactive products, such as honey or inert dressings. The FDA offers clear guidance in the distinction between drugs and devices.³³ Put simply, a medical device is a treatment that meets the criteria for drugs but “does not achieve its primary intended purposes through chemical action… and which is not dependent upon being metabolized,” making it exempt from the more rigorous requirements for drug approval.³³

It is important to understand these differences between treatments, and, most importantly, the magnitude of difficulty in developing, testing, and launching a new drug, which is exponentially more challenging and can cost hundreds of millions of dollars to achieve.³⁴ These drastically different barriers to entry explain, in part, the relative abundance of non-drug wound care therapies versus novel biologically active drugs.

Although these non-drug products are extremely valuable options, they do not directly target deficits in diabetic skin or the pathophysiology of DFUs, and additional specific and biologically active treatments that work to rescue aberrant healing are necessary. Understanding why barriers to novel drug development exist requires an understanding of the models, methods, and major advances in diabetic wound healing research.

Only then can scientists begin to answer these key questions: What is still missing? How can approaches be updated for novel therapeutic development? Which discoveries can be expanded to surmount the challenges diabetic wound healing presents? Also most importantly, how can sufficient clinical advancement be achieved to protect the growing patient population with life-threatening complications of DM?

Discussion

Common murine models of DM

Experimental studies in animal models of DM began centuries ago, with Brunner's early description of polyuria, polydipsia, and polyphagia in pancreatectomized dogs in 1722, Minkowski and von Mering's pancreatectomy-induced canine diabetes model in 1889, and Banting and Best's isolation of insulin in 1921, which led to the life-saving development of exogenous insulin therapy for humans.^35,36 Today, many model systems exist to enable investigations into both type 1 (DM1) and type 2 diabetes (DM2), each with their own drawbacks and advantages (Table 1).^37–55 The vast majority of these are rodents, categorized by whether they are genetic, inducible, or spontaneous models of DM1 or DM2 (Fig. 4).⁵⁶

Table 1.

Comparison of common rodent models of diabetes mellitus

Model+TypeReferences	Description	Onset of Hyperglycemia	Wound Healing Impairment
			2–5 daysa	6–10 daysa	11–20 daysa
Streptozotocin-induced: Inducible, DM138–40	These mice are dosed with either streptozotocin, a bacterial antibiotic, or alloxan, a pyrimidine derivative similar in structure to glucose. Onset and severity vary based on whether treatments are a single high dose or multiple low doses. Both substances cause destruction of the pancreatic beta cells, mimicking DM1.	Reaches permanent diabetes within days, pattern varies in single vs. multiple dose models	Single dose: 1.53a (95% CI 0.96 to 2.10) Multidose: 1.16a (95% CI 0.84 to 1.47)	Single dose: 2.09a (95% CI 1.35 to 2.84) Multidose: 1.69a (95% CI 1.27 to 2.11)	Single dose: 3.58a (95% CI 2.57 to 4.59) Multidose: 1.53a (95% CI 1.06 to 1.99)
Alloxan-induced: Inducible, DM139,40			Single dose: 1.20a (95% CI 0.42 to 1.99) Multidose: 5.35a (95% CI 2.50 to 8.20)	Single dose: 2.32a (95% CI 0.63 to 4.01) Multidose: 5.89a (95% CI 2.80 to 8.98)	Single dose: 2.05a (95% CI 0.38 to 3.72) Multidose: 5.84a (95% CI 2.77 to 8.91)
Nonobese diabetic: Spontaneous, DM139,41,42	The NOD mouse is a spontaneous autoimmune model of diabetes originating from the Cataract Shionogi mouse. It mimics DM1, and affected mice have some shared autoantibodies with humans.	Develops by 30 weeks, higher incidence in female mice	1.47a (95% CI 0.68 to 2.27)	6.09a (95% CI 4.47 to 7.72)	14.59a (95% CI 4.46 to 24.71)
High-fat diet: Inducible, DM239,43–45	These mice are fed a high-fat diet for a specified time frame to mimic obesity, metabolic syndrome, and DM2. Presentation and onset varies widely based on the diet, duration, and mouse strain.	Severe obesity and hyperglycemia by 16 weeks; varies with diet/strain	0.17a (95% CI −0.89 to 1.23)	1.16a (95% CI 0.38 to 1.95)	1.34a (95% CI 0.65 to 2.03)
Ob/Ob: Genetic, DM239,42,47	These mice are homozygous mutants for the Lep gene, which encodes Leptin, an adipose hormone involved in hunger regulation. These mice mimic human metabolic syndrome and are severely obese. Onset and duration of hyperglycemia varies by background.	Onset at around 4 weeks, transient (up to 14–16 weeks) in C57BL/6J	3.55a (95% CI −0.07 to 7.18)	4.57a (95% CI 2.1 to 7.04)	—
Db/Db: Genetic, DM239,42,46	These mice are homozygous mutants for the Lepr gene, the receptor for the Leptin hormone. They mimic human DM2 and exhibit very similar disease onset, progression, presentation, and severity. Db/Db mice are the leading animal model of DM2.	Hyperinsulinemic by 10–14 days, hyperglycemic by 4–8 weeks in C57BLKS/J	1.18a (95% CI 0.76 to 1.6)	2.9a (95% CI 2.29 to 3.52)	3.87a (95% CI 3.11 to 4.63)
TallyHo mouse: Genetic, DM249,57	The TallyHo mouse is polygenic model of DM2, which spontaneously develops diabetes similarly to humans. Its phenotype is less severe than other genetic models like the Db/Db mouse, and it is a more representative model of the complex inheritance of diabetes.	Mild hyperglycemia at 6–8 weeks, severe by 14–16 weeks; usually males only	Known wound healing deficits in incisional, excisional, and ischemic-reperfusion wound models.
New Zealand Obese: Genetic, DM252,53	The NZO mouse is a polygenic model of severe obesity that is highly similar to human metabolic syndrome and, in male mice, can progress to mature onset DM2.	Late onset at 20–24 weeks, in males; can be accelerated with high-fat diet	Not commonly used in wound healing studies.
Otsuka Long-Evans Tokushima Fatty Rat: Genetic, DM251,55	The OLETF rat is an inbred polygenic model of chronic DM2 with mild obesity and late onset hyperglycemia. Like many of the other genetic models of DM, primarily males develop full-blown disease.	Late onset after 18 weeks; usually males only	Known impairments in corneal wound healing.
Zucker diabetic fatty rat: Genetic, DM248,50,54	ZDF rats are derived from select homozygous fa/fa mutants of the original Zucker fatty rat strain. These rats develop early obesity and insulin resistance which mimics the progression to DM2 in humans. Interestingly, most female ZDF rats do not become hyperglycemic.	Mild hyperglycemia at 7–8 weeks, severe by 14–16 weeks; usually males only	Impaired wound closure, abnormalities in wound tissue structure, inflammation, and scarring.

^aValues represent mean difference in wound size compared to nondiabetic controls, from 2020 meta-analysis of diabetic ulcer models.³⁹

Days, days postwounding; DM, diabetes mellitus; NOD, nonobese diabetic; NZO, New Zealand Obese; OLETF, Otsuka Long-Evans Tokushima Fatty; ZDF, Zucker diabetic fatty.

Figure 4.

Common rodent models of DM. An overview of some of the most common murine models of type 1 and type 2 DM. For type 1 DM, these include the streptozotocin and/or alloxan-induced models and the nonobese diabetic mouse.^37,56 For type 2 DM, they include the high-fat diet mouse, leptin and leptin receptor mutants, Zucker diabetic fatty rats, and polygenic models such as the New Zealand Obese mouse and the Otsuka Long-Evans Tokushima Fatty Rat.^10,11,57 DM, diabetes mellitus. Color images are available online.

The most commonly used models are those that best approximate the human pathogenesis and development of DM. For DM1, these include the streptozotocin- or alloxan-induced mice and nonobese diabetic mice, all of which undergo destruction of the insulin-producing pancreatic beta cell, similar to the autoimmune pathogenesis of DM1 in human patients.^37,56 For DM2, which affects the vast majority of diabetic patients, the most prominent models include the genetic leptin mutant ob/ob and leptin receptor mutant db/db mice, Zucker diabetic fatty rats, high-fat-diet-induced mice and rats, and polygenic models such as the Otsuka Long-Evans Tokushima Fatty rats, TallyHo mice, and New Zealand Obese mice.^10,11,57

Because murine models remain the most widely studied, they comprise the majority of diabetic wound healing research. However, larger animals, including swine models such as the Ossabaw and Yorkshire pigs, are emerging as model organisms that come increasingly close to mimicking the healing properties of human skin.^58–60 When compared to humans, porcine models demonstrate more analogous skin structure, increased concordance (78%) in study results, and more similar responses to treatments than both small mammal (53%) and in vitro (57%) models.^58,61

As these more clinically relevant models become increasingly well-characterized, they warrant significant discussion as potential solutions to this problem. Although they present a series of practical considerations, including increased costs and experimental difficulty, the increased concordance between human and porcine models could be key to limiting the number of laboratory discoveries that fall short during clinical trials.

Alongside the animal models of DM, there is a wide array of experimental methods used for the study of wound healing (Fig. 5). Among the in vivo assays, the excisional wound model is the most prominent. In this model, a full-thickness skin segment is removed to enable investigation of wound closure, reepithelialization, cell migration and prolif eration, angiogenesis, formation of granulation tissue, scarring, and more.⁶² Other commonly used methods include partial thickness wounds, incisional wounds, ischemic cutaneous wounds, surface burn wounds, and abrasion and tape-stripping wounds,⁶² each of which serves its own utility in investigating the wound healing process and its aberrancy in DM.

Figure 5.

Common experimental methods of wound healing. A summary depicting some of the most commonly used experimental wounding models, including excisional wounds, incisional wounds, ischemic wounds, burn wounds, and abrasion wounds.^62,63 Color images are available online.

In addition to these in vivo methods, there are many standard in vitro protocols that are regularly utilized, especially in the early stages of research. These include 3D cultures, co-cultures, single-cell studies, organoids, scratch and migration assays, and more.⁶³ Because the majority of drug development studies progress from these models into animal models before reaching human subjects, in vitro methods are not discussed in depth here. They do, however, play an essential role in wound healing research and are often a key starting point for any potential therapeutic.

Due in no small part to the wide variety of models and methods available, both the quantity and quality of diabetic wound healing research has increased exponentially in recent decades, leading to immense discovery of the individual components that play a role in the development and healing of diabetic ulcers. A PubMed search for “diabetic wound healing” yields only 3 results for the year 1962, but nearly 900 citations appear in the results for the year 2019. Massive growth in this area has brought with it many achievements in the murine models above, suggesting various biological pathways, compounds, cell types, growth factors, small molecules, cytokines, and more that have improved diabetic wound healing experimentally.

Here, we summarize many of these discoveries in murine models and discuss the challenges and constraints that impact their translational success. Of note, the breadth of diabetic wound healing research extends far beyond these direct, biologically derived or bioactive factors and includes other approaches such as novel and supplemented methods of wound care, novel wound dressings and drug delivery methods, gene therapy, adaptations of existing FDA-approved medications, and more, which cannot be discussed in-depth in this review.^28,64–67

Growth factors

Among the most prominent and well-studied discoveries for the improvement of diabetic wound healing is the application of growth factors.^68–70 Of these, the most successful is PDGF-BB, which was first shown to improve diabetic wound healing in db/db mice in the early 1990s and reached Phase I clinical trials by 2000.^14,71–76 To date, Regranex (becaplermin gel), the topical recombinant PDGF-BB therapy, remains the only FDA-approved prescription drug specifically for DFUs, although its usability, practicality, and effectiveness in the clinic are relatively poor despite significant efficacy in clinical trials.^12,14,16,77

Additional research continues to investigate the effects of PDGF on diabetic wound healing, including in combination with other growth factors, such as fibroblast growth factor (FGF),^71,78 vascular endothelial growth factor (VEGF),⁷⁹ and transforming growth factor alpha (TGFα).⁷² Studies in diabetic mice and other healing-impaired models demonstrated that recombinant acidic or basic FGF each improved tissue repair, dermal wound healing, and angiogenesis as early as the 1990s.^80–85

Several more recent studies, beginning in 2011, also demonstrated the beneficial effects of FGF on diabetic wound healing using rat models and novel polygenic mouse models; efficacy was also observed in studies using other experimental methods, such as gene transfer, joint applications with other factors, modified applications of FGF, and novel drug delivery vehicles, such as electrospun fibers and fibrin scaffolds.^86–91

A similar pattern exists for investigations involving VEGF. Beginning in the early 2000s, Rico et al. and Galiano et al. demonstrated that topical application of VEGF to dermal wounds in db/db mice significantly sped wound closure and improved wound healing.^92,93 Additional studies over the next 15 years continued to demonstrate the effectiveness of VEGF therapy in diabetic wound healing models, using methods and tools such as gene therapy, combination therapy with other factors, collagen and fibrin scaffolding, nanospheres, and more.^{79,86,89,93–96}

Others' investigations have also demonstrated the effectiveness of many other growth factors in improving diabetic wound healing experimentally, including keratinocyte growth factor (KGF),^97–99 epidermal growth factor (EGF),^95,100–105 TGFα and TGFβ,^72,106–110 insulin-like growth factor 1,^{107,111–115} connective tissue growth factor,¹¹⁶ hepatocyte growth factor,^117,118 and nerve growth factor.^119,120

These discoveries provide an important opportunity to discuss diabetic wound healing research beyond preclinical investigation in model organisms, when a promising experimental treatment reaches human patients and, often unexpectedly, fails during late-stage clinical trials. Growth factors, in particular, represent a large portion of this group, making them an excellent example of the disparity between the laboratory and the clinic and giving a glimpse into where the biggest challenges lie.

Cellular therapies

Another major category yielding experimental success in diabetic murine models is the application of specific cell populations to improve wound healing. These include stem and progenitor cells, stromal cells, immune cells, fibroblasts, and keratinocytes. Among the stem and progenitor cell populations, mesenchymal stem cells (MSCs) are the most extensively studied and have been shown to improve diabetic wound healing in various applications.^121–130 The earliest of these studies began in 2007 and used MSCs derived from autologous bone marrow, which improved wound healing both in db/db mice and in a small number of human patients with acute and chronic wounds.^121,124

Since then, further preclinical investigations have demonstrated positive effects on wound healing in response to adipose-derived MSCs,^{125,129,131–142} umbilical cord blood-derived stem cells,^{123,143–150} embryonic stem cells,^151,152 hematopoietic stem cells,¹⁵³ endothelial progenitor cells (EPCs),^{147,154–157} and CD34⁺ cell populations.^{143,150,158–161} In addition, stromal cells derived from both bone marrow^162,163 and adipose tissue,^164–166 immune cell populations such as mac rophages¹⁶⁷ and mature B cells,^167,168 fibroblasts,^148,169 keratinocytes,¹⁷⁰ and platelets^171,172 have all been shown to improve wound healing in murine models of DM.

Each of these cell-based therapies demonstrated preclinical efficacy; however, the majority never reached clinical trials, and even fewer emerged successful. One such example is the macrophage-based therapeutic CureXcell by MacroCure Ltd. The application of macrophages to wounds demonstrated early efficacy in human patients in 2004, leading to the development of MacroCure's namesake product.¹⁷³ CureXcell, a then cutting-edge biologic, was on track to become one of the few diabetic wound care options on the market, until it failed in Phase III clinical trials,^173–177 serving as an example of lost efficacy when study scale increases and practicality becomes a critical metric.

Much like macrophages, platelet therapies for wound healing have also undergone significant investigation in human patients, including multiple clinical trials; several variations, including platelet-rich plasma, platelet-rich fibrin, and autologous platelet gel, are in use clinically for diabetic and other wounds, and further trials are ongoing.^178–184

Of the many cell-based advances mentioned here, the most prominent and enduring clinical success is the use of fibroblasts and/or keratinocytes to create HSEs that can be applied with appropriate dressings to improve healing of chronic DFUs. Two such HSEs, Apligraf (Graftskin),^21,22,24,185 and Dermagraft,^17,186 are FDA-approved for DFUs lasting more than 4–8 weeks, although they cannot be used in with standard topical wound care agents or for any DFU with infection or other common complications. It is of important note that Apligraf, Dermagraft, and their drug counterpart, Regranex, were all approved around the turn of the century: Regranex in 1997, Apligraf in 1998, and Dermagraft in 2001.^13,18,23

In the 20 years since, Integra's Omnigraft Dermal Regeneration Matrix, an animal-derived wound dressing containing bovine collagen, shark cartilage, and silicone, originally approved for burns in 1996, gained additional approval for use in diabetic wounds in 2016.^27,187 Considering the exponential growth of diabetic wound healing research over the same 20-year period, this statistic underscores a somber scientific reality: although diabetic wound healing research models and methods have been established that produce valuable laboratory results, there clearly remains a knowledge gap impeding the clinical progress they ultimately seek to achieve.

Cytokines, chemokines, and other cellular activation molecules

In addition to harnessing the potential of whole-cell populations, many have investigated the use of individual cytokines, cellular proteins, and signaling molecules to improve diabetic wound healing.¹⁵⁵ These include IL2,^188,189 IL4,¹⁹⁰ IL6,^191,192 IL22,¹⁹³ TNFα,¹⁶⁷ and CCL2,^194,195 all of which contribute to improved wound repair in diabetic murine models.

Many other endogenous proteins or protein an alogs have also demonstrated contributions to diabetic wound healing, including erythropoietin,^{105,130,196,197} substance P,^198–203 in sulin,^103,204,205 C peptide,^206,207 nuclear factor erythroid 2-related factor 2 (NRF2),^208–212 Sonic Hedgehog,²¹³ homeobox protein HoxD3,²¹⁴ MMP-9,^215–217 oncostatin M,²¹⁸ nesfatin-1,²¹⁹ thyroid hormone,²²⁰ mineralocorticoid receptor antagonists,²²¹ GnRH antagonists,²²² and inositol-requiring enzyme 1.²²³

One prominent example of a cellular activation-based therapy is the topical application of platelet releasate (PR), also known as platelet-derived wound healing factors/formula, to treat nonhealing wounds. PR is a mixture of platelet-released proteins, chemokines, and cytokines that is usually derived from autologous platelets and first demonstrated improved wound healing in humans in 1986.²²⁴ Randomized human trials using PR began in 1990,^225–227 and additional research demonstrated the effectiveness of PR for wound healing in several animal models, including diabetic mice and rats.^228–231

Over the next two decades, investigation continued into the use of PR for DFUs and other nonhealing wounds, and many clinics today offer autologous PR as a treatment option. While PR and similar substances, such as platelet-rich plasma and autologous platelet gel, are not FDA-approved drugs, their clinical use continues to increase, and further investigations are ongoing.^231–234

While the majority of studies demonstrating improvements in diabetic wound healing do so using single factors, PR and its more recent counterparts, such as transcription factors, microRNAs (miRNAs), and exosomes, represent potential therapeutics that may be able to rescue multiple inherent shortcomings in the diabetic wound at once, thereby improving wound healing more significantly than any single factor could alone.

PR serves as a very early example of these broader investigations, a key feature that may explain the longevity of PR as a research target for wound healing and its ongoing clinical use. Expansion into more widely efficacious therapies is rapidly increasing and marks an important step in recognizing the limitations of research discoveries thus far and adapting the investigational approach.

Alongside proteins and cytokines, a multitude of chemical compounds and small molecules have demonstrated experimental efficacy in improving diabetic wound healing. Of the compounds studied, carnosine,²³⁵ quadrol,²³⁶ sphingosine 1-phosphate,²³⁷ deferoxamine,^238–240 nonadenine-based purines,²⁴¹ fullerenes,²⁴² the DPP-4 inhibitor MK0626,²⁴³ ND-336,²⁴⁴ hydroxylase inhibitors,²⁴⁵ and hydrogen sulfide^246–248 have all been shown to accelerate wound healing in murine models of DM2.

Many studies not discussed here have also investigated the potential repurposing of existing FDA-approved medications for other conditions to be used for DFUs, such as beta-blockers, anesthetics, metformin, and others. Drug repurposing increases the likelihood that clinical trials may be successful since they rely on substances that have already demonstrated safety and efficacy for at least one indication in human patients, making this method an appealing and potentially productive approach for diabetic wound healing and other conditions, including cancers, neurological disorders, and more.^249,250

However, no existing DM drugs have been reapproved by the FDA for diabetic wound care, although there are off-label uses. This limitation likely comes on the heels of a few key obstacles: lack of interest from pharmaceutical companies; insufficient funding for the still very costly trials, even for a previously approved compound; and limited access to and strict regulation of existing drugs (especially those for which affordable generics do not yet exist).^251,252 These limitations hinder the potential of a vast library of existing and likely underutilized molecules, many of which could have additional therapeutic possibilities.

One such example is Galnobax, a topical DFU treatment containing esmolol hydrochloride, a beta blocker commonly used as an antiarrhythmic, for which recruitment began for a Phase III clinical trial in June of 2019.²⁵³ Although the study's outcome is still pending, Galnobax represents an important foray into the repurposing of existing drugs for the treatment of DFUs. Time will tell, however, whether the costly and intricate ventures of a clinical trial, especially during Phase III, can pay off using “old” drugs.

MicroRNAs

As the understanding of cellular functions has deepened, diabetic wound healing research has progressed to previously impossible targets, such as miRNAs. MiRNAs are short noncoding RNA sequences that modulate gene expression by directly targeting matching mRNA sequences to control a wide breadth of cellular functions.²⁵⁴ Diabetic wound healing research has begun to investigate at the miRNA level to understand the effects of DM on miRNA expression during wound healing and the ways in which miRNAs could be modulated to rescue impaired healing. In a single miR profiling study in diabetic rats in 2015, Liu et al. revealed 83 distinct miRNAs that are differentially expressed in DFUs.²⁵⁵

Several individual miRNAs have been studied to further understand their contributions. MiR-129-2-3p,²⁵⁶ miR-132,²⁵⁷ miR-129 and miR-335,²⁵⁸ and miR-27b²⁵⁹ have all been shown to directly improve diabetic wound healing. Others, including miR- 146a,^260,261 miR-15b,^262–264 miR-200b, miR-126,^265,266 and miR-21,²⁶⁷ are known to be involved, although their contributions and therapeutic potential remain less well-understood.

Some miRNAs have been shown to contribute to impaired wound healing in DM, making them good targets for inhibitory therapy. MiR-155 inhibition,^268,269 miR-15b and miR-200b inhibition,²⁶⁴ the modified miR-92a inhibitor MRG-110,²⁷⁰ and the miR-26a inhibitor LNA-anti-miR-26a²⁷¹ improve cutaneous wound healing in murine models of DM.

Importantly, investigation into the roles of miRNAs in diabetic wound healing also requires consideration of epigenetic regulation, which both alters gene expression levels and affects protein translation. Differences in epigenetic patterns have been observed in human patients with DM and other conditions,²⁷² and both miR-200b and miR-19a have been implicated in the regulation of vascular and tissue factors in DM.^273,274 These and other epigenetic targets are a young area of diabetic wound healing research with important advantages.²⁷⁵

MiRNAs and epigenetic regulation can readily influence gene expression within a cell, drastically expanding their potential for efficacy over a single soluble factor. Because of this increased breadth, they represent an important level of innovation that could potentially overcome the deficits that single factors have so far been unable to conquer.

Microvesicles and exosomes

Similarly, recent diabetic wound healing studies beginning in 2015 have turned to extracellular vesicles as potential therapeutics. Both microvesicles and exosomes are specific types of extracellular membrane vesicles that carry packaged proteins, RNAs, signaling molecules, and other substances from cell to cell, playing a crucial role in communication and regulation.²⁷⁶ Because these biological packages are capable of carrying a heterogenous combination of factors and information from their parent cells, they are more broad-spectrum research targets that could inform the understanding of diabetic wound healing with greater resolution.

Exosomes from cell populations, including fibrocytes,²⁷⁷ EPCs,^156,157 platelet-rich plasma,^278,279 MSCs,^280,281 and adipose stem cells,²¹¹ have all demonstrated efficacy in diabetic mouse and rat models. Like their miRNA counterparts, microvesicles offer the potential promise of much more widespread targeting within the diabetic wound, with their contents representing an amalgam of factors that are, together, more potent and diverse in their effects.

Bio- and bioengineered materials

Another state-of-the-art research approach has become increasingly more common over the last 15 years: the use of bio- and bioengineered materials to modulate wound repair.⁶⁶ These efforts aim to modernize and expand the boundaries of diabetic wound healing research by taking advantage of rapid technological advancement. Many of these studies have used biomaterials to elaborate on previously studied factors, making unique modifications to improve the target's potential therapeutic effects.

Such examples include biomaterial-grown EPCs,²⁸² electrospun biomaterial seeded with MSCs,²⁸³ drug-eluting bioengineered scaffolds containing macrophage chemokines,²⁸⁴ adipose stem cells within an artificial dermis,¹³⁴ EGF-releasing hydrogel,²⁸⁵ and more. Still others of these more tech-savvy approaches have produced a rapidly growing body of preclinical research demonstrating promising potential for diabetic wound care, including a variety of hydrogels,^{141,286–302} nanoparticles,^{89,261,303–309} and scaffolds.^{89,96,290,300,305,310–316}

Here, it is critical to note the many other mechanical, physical, and external wound-modulating treatments, such as LED light therapy,^317–321 vacuum-assisted wound closure,^322–326 a wide variety of dressings and bandages, and many more. Although they are generally not bioactive, these options are also rapidly evolving, due to the same technological advances that drive the new wave of bioengineered and biomaterial solutions. Together, these studies represent an important but still-young shift in the approach to diabetic wound healing research, pushing beyond the traditional low-tech approaches that comprise the majority of the existing body of work.

As bioengineered and more tech-driven methods continue to expand, they may be able to clear the hurdles that have delayed clinical progress thus far and achieve previously impossible efficacy in driving wound repair.

Plant- and animal-derived substances

By far the largest and most diverse group of factors that have been suggested to improve diabetic wound healing are plant- and animal-derived substances, including readily accessible natural products, herbal treatments, ingredients commonly included in skincare products, and substances found in our everyday diet. The most common and well-studied of these are curcumin^{199,308,327–330} and curcumol,³³¹ bee venom,^210,332,333 honey,^334–337 propolis,^338,339 resveratrol,^340,341 aloe vera,^342,343 and whey protein.^344–346

Other common, although less thoroughly investigated, substances that have been shown to improve diabetic wound healing include nicotine,³⁴⁷ vitamin D,³⁴⁸ l-arginine,³⁴⁹ yeast extract,³⁵⁰ citrus peel extract,³⁵¹ Annona squamosa (sugar apple),^352,353 Rosmarinus (rosemary),³⁵⁴ and spinach extract.³⁵⁵ There are several animal-derived substances that have also demonstrated some therapeutic potential, including frog skin peptides,³⁵⁶ chum salmon skin gelatin,³⁵⁷ acellular fish skin,^358,359 powdered shells of the snail Megalobulimus lopesi,³⁶⁰ and camel milk peptides,³⁶¹ as well two funguses, Sparassis crispa (cauliflower mushroom),^362,363 and proteins from Phellinus gilvus (mustard yellow polypore mushroom).³⁶⁴

The majority of the other substances studied are derived from plants, including aescin (from horse chestnut),³⁶⁵ Martynia annua (cat's claw),³⁶⁶ Lithospermum,³⁶⁷ Catharanthus roseus (periwinkle),³⁶⁸ Momordica charantia (bitter gourd/melon),^369,370 pi per betel,³⁷¹ the herbal medicine “Yi Bu A Jie”,³⁷² Mallotus philippinensis (kamala tree),³⁷³ Cotinus coggygria (smoke tree),³⁷⁴ Stachys hissarica,³⁷⁵ Angelica dahurica,³⁷⁶ Moringa oleifera,³⁷⁷ Rehmanniae radix,^378,379 Lycium depressum,³⁸⁰ Morinda citrifolia (noni) juice,³⁸¹ Carica papaya,³⁸² and Strychnos pseudoquina extract.³⁸³ It is of note that some of these substances, most especially honey, have long been incorporated into standard wound care products, such as honey dressings, bandages, and topical gels and creams, which have proven clinically successful for general wound care.³⁸⁴

Clinical trials

Although more and more potential therapies are being developed in small mammal models of DM, only a few go on to large animal testing, and even fewer still make it through clinical trials to achieve FDA approval. This creates a costly and disheartening stream of failed trials, marking the most critical bottleneck in the development of effective, practical treatments for DFUs.

Among the most notable downfalls are CureXcell, the macrophage-based topical therapy discussed previously, which failed its Phase III clinical trial in 2015^174,177; Trafermin, also known as Fiblast, a recombinant human basic FGF spray for DFUs, which failed its Phase III clinical trial in 2014^385,386; DermaPro, a topical diperoxochloric acid, which failed a phase III clinical trial in 2015^387,388; Aclerastide, a gel containing the compound DSC127, for which the Phase III trial was terminated during recruitment due to demonstrated futility of the drug^389–391; Locilex (pexiganan), a topical cream containing a 22-amino acid polypeptide to treat infected DFUs, which failed two Phase III clinical trials in 2016^388,392,393; and Repifermin, a topical treatment containing KGF2, which failed a Phase II clinical trial in 2003.^394,395

These highly publicized attempts represent only a small fraction of the total number of clinical trials for DFUs. As of early 2020, a search for “diabetic foot ulcer” trials in the ClinicalTrials.gov database reveals 63 Phase I trials, 105 Phase II trials, and 73 Phase III trials—a total of 196, of which only 22 have listed results.³⁹⁶ Of the 196 trials, 96 are listed as completed, 26 are listed as terminated, suspended, or withdrawn, 29 are in the prerecruitment, recruitment, or invitation phases, and only 7 are currently active (leaving the remaining 38 studies with status unknown).³⁹⁶ While the ongoing recruitment and active trials offer great promise for new treatments, these numbers also hint at the bigger picture: a massive amount of progress that often yields little to no clinically usable result by Phase III.

The factors driving this high rate of late stage attrition are not limited to the shortcomings of the animal models and the limited scope of research targets. Arguably one of the most critical hurdles is inherent to the clinical trials themselves: the simplicity of the trial cohort compared to the highly complex and diverse patient population. To avoid confounding variables and ensure reliable results, clinical trials must be designed with strict and specific inclusion and exclusion criteria. This aspect of study design, while essential for high-quality data, introduces its own dilemma: if the patient population does not match the study cohort, how generalizable are the results?

In the case of patients with DFUs, this is a common problem. Take Regranex, for example: in clinical trials, it was tested only in diabetic patients with neuropathic ulcers of a limited size that lacked any other complicating factors, such as vasculopathy or infection. The general population of patients with DFUs, however, includes large numbers of patients with peripheral vasculopathies, wound complications, and many additional comorbidities that could all reduce the effectiveness of any treatment. This added complexity likely explains the difference between Regranex's impressive success in clinical trials compared to its more modest Phase IV results.³⁹⁷

For therapies deemed ineffective in Phase III, this issue may simply be coming to light earlier on, when the increase in cohort size introduces enough variability to eliminate significant results. When considered alongside the limitations of preclinical research, the complexity of the patient population is of great consequence. While it would be entirely unfeasible to conduct clinical trials large and varied enough to fully account for patient diversity, it must be considered during trial design to maximize the impact of the study results.

Bench-to-bedside barrier

The individual research targets discussed here (summarized in Table 2) are only a small subset of the many hundreds of studies that have unveiled potential therapeutic targets for diabetic wound healing. Countless other types of advances exist in murine models of DM, including discoveries in wound care, drug repurposing, light-based therapies, novel application methods, genetic editing, retroviral studies, and much more. In total, there are significantly more potential advances than any one review can encompass. Yet, the clinical achievements derived from all of these discoveries can be easily summarized, highlighting the incredible bottleneck between the bench and the bedside.

Table 2.

Murine diabetic wound healing research targets

Growth Factors		Cellular Therapies	Microvesicles and Exosomes From	Bio- and Bioengineered Materials
PDGF-BB	EGF	Stem/Progenitor cells (MSCs, adipose-derived cells, umbilical/cord blood stem cells, HSCs, CD34+ cells, EPCs)	Fibrocytes	Biogrown EPCs	LED light
FGF	IGF-1	Stromal cells	EPCs	Electrospun fibers	Vacuum-assisted closure
VEGF	CTGF	Immune cells (macrophages, monocytes, B cells)	Platelet-rich plasma	Drug-eluting scaffolds	Negative pressure therapy
TGFα, TGFβ	HGF	Fibroblasts	MSCs	Hydrogels	DermaPACE
		Keratinocytes	Adipose stem cells	Nanoparticles	Engineered dressings
		Platelet-rich plasma, platelet gel

Plant- and Animal-Derived Substances			Cytokines, Chemokines, and Signaling Molecules			MicroRNAs
Curcumin	Powdered snail shell (Megalobulimus lopesi)	Carica papaya	IL2	MMP-9	Nonadenine purines	miR-129-2-3p
Curcumol	Camel milk peptides	Martynia annua (cat's claw)	IL4	Oncostatin	Fullerenes	miR-132
Bee venom	Stachys hissarica	Lithospermum	IL6	Nesfatin-1	Hydrogen sulfide	miR-129
Honey	Phellinus gilvus (mustard yellow polypore mushroom)	Catharanthus roseus (periwinkle)	IL22	Thyroid hormone	β-Blockers	miR-335
Propolis	Chum salmon skin gelatin	Momordica charantia (bitter gourd/melon)	TNFα, TNFβ	Mineralocorticoid receptor agonists	Anesthetics	miR-27b
Piper betel	Yeast extract	“Yi Bu A Jie” (herbal medicine)	CCL2	GnRH antagonists	Metformin	miR-146a
Resveratrol	Citrus peel extract	Mallotus philippinensis (kamala tree)	EPO	Inositol-requiring enzyme	Sonic hedgehog	miR-15b
Aloe vera	Spinach extract	Angelica dahurica	Substance P	Carnosine	HoxD3	miR-21
Whey protein	Rosmarinus (rosemary)	Sparassis crispa (caulifolower mushroom)	Quadrol	Insulin, C peptide	DPP-4 inhibitors	miR-155 inhibitors
Nicotine	Annona squamosa (sugar apple)	Cotinus coggygria (smoke tree)	NRF2	Deferoxamine	ND-336	miR-92a inhibitor (MRG-110)
Vitamin D	Frog skin peptides	Moringa oleifera	Platelet releasate	Sphingosine 1-phosphate	Hydroxylase inhibitors	miR-26a inhibitor (LNA-anti-miR-26a)
l-Arginine	Lycium depressum	Rehmanniae radix
Aescin	Morinda citrifolia (noni) juice	Strychnos pseudoquina

CTGF, connective tissue growth factor; EGF, epidermal growth factor; EPC, endothelial progenitor cell; EPO, erythropoietin; FGF, fibroblast growth factor; HGF, hepatocyte growth factor; HSC, hematopoietic stem cell; IGF-1, insulin-like growth factor 1; MMP-9, matrix metalloproteinase 9; MSC, mesenchymal stem cell; NRF2, nuclear factor erythroid 2-related factor 2; PDGF, platelet-derived growth factor; TGF, transforming growth factor; VEGF, vascular endothelial growth factor.

While the factors contributing to this limited progress are not fully understood, three noteworthy contributors stand out. The first is the relative insufficiency of the murine models of diabetic wound healing. Despite their variety, they all fall short of one of the most critical features of the diabetic wound: a chronic inability to heal. In the commonly studied murine models of DM, experimentally induced wounds do heal independently, although more slowly than their wild-type counterparts, and they do not develop chronic, nonhealing wounds without significant additional interventions such as induced infection or repetitive injury.

Furthermore, animal models of DM do not spontaneously develop ulcers. In every case, wounds must be created or induced to investigate their healing, meaning that (a) the wounds are acute, rather than chronic, and (b) they do not represent the insidious, involuntary, and often initially unnoticed development of ulcers seen in the diabetic patient. These caveats suggest that, fundamentally, the prominent animal models of DM do not sufficiently recapitulate the human condition for DFUs.⁶¹ Therefore, the many scientific discoveries that show immense promise in these models can easily fall short when translated to humans, likely because their contribution is enough to speed the healing of an acute, slow-healing wound but not enough to rescue one that remains chronically unable to heal.

The second notable contributor is the repeated pattern of investigating single factors in the diabetic wound. Of course, this is not to minimize the value of specific, mechanistic investigations. These studies have been invaluable in their contribution to the understanding of the wound environment and have uncovered critical details about the aberrant processes and inherent deficits in the diabetic wound. Furthermore, they have proven critical in enabling mechanistic insights and a more functional understanding of the pathogenesis of chronically nonhealing diabetic ulcers. However, single-factor studies have a major shortcoming in terms of clinical translation: insufficient power to correct the level of aberrancy in the human diabetic wound.

Despite their value at the basic science level, each individual small molecule, extract, or compound has a very narrow scope of effect. This level of specificity renders each factor highly informative relative to its own targets, but leaves much to be desired in the grand scheme of the wound environment. As a consequence, the effect size of each factor becomes relatively small and is almost always insufficient to rescue the severe healing impairment observed in human DFUs.

Finally, and relatedly, the third major contributor is the clinical diversity of the patient population. Because of the large of number of complications and comorbidities among patients with DM, clinical trial cohorts cannot perfectly represent the population and must rely on more specific subgroups to obtain scientifically valuable results. When therapies go on to reach larger patient populations, where those complicating factors are reintroduced, the efficacy of the new treatment is often greatly reduced.

When considered in combination, these major factors reveal an important paradigm regarding diabetic wound healing research: because the prominent models do not fully recapitulate the human condition, the majority of investigations center on single factors affecting healing, and study cohorts cannot replicate the diversity of the patient population, an insufficient research approach for the development of therapeutics for chronic diabetic wounds has been established.

To address this growing and unmet need, scientists must work to overcome these obstacles by expanding and adapting the research methodology. Novel approaches must be considered, including the development and validation of more specific and relevant model systems, expanded clinical trial populations, multifaceted research strategies that view the diabetic wound through a wider lens, and layered innovations into the management of diabetic wounds, such as combination therapeutics and preventative measures.

Of course, many of these efforts are not without their own challenges. Combination therapies in particular pose a unique set of problems: even with solid scientific foundations behind them, these potential treatments face hurdles in regulation and testing, higher costs, licensing and patent difficulties, and complex battles in garnering the interest and cooperation of the parties involved.³⁹⁸ In some cases, these challenges may prove insurmountable, but existing combination therapies for conditions such as HIV infection and cancer offer hope that they are not entirely impossible.

Cutting-edge advances across disciplines, including medicine, technology, and mathematics, must also be used to empower forward progress, such as the use of complex computational models, identification of new biomarkers, and investigation into the possibilities of personalized medicine for wound care. Promisingly, many innovations are already beginning to tackle these obstacles head-on, and the rapid expansion of these new research areas suggests a bright future for diabetic wound healing therapies.

Evidence of this progress is already beginning to improve diabetic wound care. Several recent multicenter randomized controlled trials investigating the effectiveness of Grafix, LeucoPatch, and sucrose octasulfate in improving diabetic wound healing have demonstrated positive results, prompting the inclusion of new recommendations in the International Working Group in Diabetic Foot's recent guidelines, published in 2020.^399–402 These included several new interventions to improve healing of chronic DFUs, including the use of autologous leukocyte therapies, platelet and fibrin-based products, placental products, and sucrose octasulfate dressings.⁴⁰²

As these new guidelines are implemented, patients who previously had few options for care may finally receive new treatments that could prevent amputation, underscoring both the need for a variety of effective therapies and the significant benefits of the latest advances in diabetic wound care.

Summary

Chronic, nonhealing diabetic wounds are a common, high-risk complication of DM that can lead to serious impairments, life-threatening infections, and amputation. To date, there are few treatment options available for patients DFUs, and the majority of their management relies on treatments for DM and standard wound care. To address this clinical need, a wide array of research investigations have sought to identify, understand, and rescue the deficits in diabetic wound healing.

These studies have revealed hundreds of potential therapeutic targets, including cell types, growth factors, cytokines, small molecules, miRNAs, extracellular vesicles, plant- and animal-based substances, and more. Despite this breadth of research success, however, nearly all of these discoveries have failed to reach human patients.

Three major factors likely contribute to this scientific bottleneck. The first is the comparative insufficiency of the animal models of DM, which do not develop or heal wounds in the same manner as human patients; the second is the relatively narrow scope of diabetic wound healing research, which tends to focus on individual components in healing; and the third is the sheer complexity of the patient population. Together, these factors create an environment in which a large body of valuable scientific work yields a more detailed understanding of diabetic wound healing, but produces limited therapies to manage it.

To surmount these challenges, scientists must work to develop more accurate model systems, to further expand into multifactorial investigations, and to consider more multifaceted and compound approaches to diabetic wound care. Important examples of this much-needed progress are already rising to the forefront, including broader research targets, such as miRNAs, transcription factors, novel biomarkers, and bioengineered materials; improved experimental model organisms, such as pigs and other large animal models; and multidisciplinary advances, such as personalized medicine and complex computational modeling.

These efforts are testament to the vast amount of progress ahead and demonstrate invaluable steps toward novel therapeutic development. Although ongoing obstacles are inevitable, scientists continue to evolve in pursuit of innovative therapies to address the critical needs of patients with chronic wounds. As the new frontiers of wound healing research continue to expand, so too will their resulting successes and, of course, the countless patients they will serve.

Take-Home Messages

• Chronic, nonhealing wounds are a serious complication of DM and lead to thousands of deaths and amputations annually.

• To date, there are four biologically active FDA-approved treatments for diabetic wounds: one drug, Regranex; two human cell devices, Dermagraft and Apligraf; and one animal-derived device, Omnigraft.

• The significant disadvantages and caveats to the available therapies and a lack of novel therapeutic development leave the vast majority of diabetic wound patients with limited treatment options.

• Significant research demonstrates the experimental effectiveness of a wide variety of factors in improving diabetic wound healing in murine models. These prominent models elucidate many viable, informative research targets, but do not sufficiently replicate the human condition.

• The majority of experimental discoveries rely on single factors, highlighting a potential shortcoming in the most common investigational approach.

• Despite the vast number of research advances, little progress has been made in successfully translating potential therapeutics into the clinic. Those that make it to clinical trials often fail in Phase III, likely due to the complexity and diversity of the patient population.

• The most commonly used models and approaches for the investigation of diabetic wound healing are largely insufficient for the development of novel therapeutics. To overcome these barriers, scientists must consider more innovative tools and approaches for diabetic wound healing research.

Abbreviations and Acronyms

CTGF

connective tissue growth factor

DFU

diabetic foot ulcer

DM

diabetes mellitus

EGF

epidermal growth factor

EPC

endothelial progenitor cell

EPO

erythropoietin

FDA

Food and Drug Administration

FGF

fibroblast growth factor

HFD

high-fat diet

HGF

hepatocyte growth factor

HSC

hematopoietic stem cell

HSE

human skin equivalent/substitute

IGF-1

insulin-like growth factor 1

KGF

keratinocyte growth factor

miR

microRNA (specific)

miRNA

microRNA

MMP

matrix metalloproteinase

MSC

mesenchymal stem cell

NGF

nerve growth factor

NIH

National Institutes of Health

NOD

nonobese diabetic

NRF2

nuclear factor erythroid 2-related factor 2, often written as NFE2L2

NZO

New Zealand Obese

OLETF

Otsuka Long-Evans Tokushima Fatty

PAD

peripheral arterial disease

PDGF

platelet-derived growth factor

PR

platelet releasate

TGF

transforming growth factor

VEGF

vascular endothelial growth factor

ZDF

Zucker diabetic fatty

Acknowledgments and Funding Sources

The authors thank the other members of the DiPietro laboratory: Junhe Shi, Anna Salapatas, Trevor Leonardo, Dr. Bruna Romano De Souza, and Dr. Wendy Cerny for their assistance during the revision of this work. They also extend special thanks to Dr. Stephanie Wu, for generously providing the clinical photographs for Fig. 1, and Dr. Veronica Haywood, for her valuable feedback. Funding is provided by NIH grants R01-GM-50875 and R01-AR069541-01A1. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Author Disclosure and Ghostwriting

No competing financial interests exist. The content of this article was expressly written by the authors listed. No ghostwriters were used.

About the Authors

May Barakat is an MD/PhD candidate in the laboratory of Dr. Luisa A. DiPietro. Her research interests are in diabetic wound healing and tissue repair and regeneration. Dr. Luisa A. DiPietro is a professor of Periodontics and the Director of the Center for Wound Repair and Regeneration at the University of Illinois at Chicago College of Dentistry. Her laboratory investigates the process of wound repair, including inflammation, angiogenesis, scar formation, mucosal and cutaneous tissue repair, and wound healing outcomes in abnormal states such as DM or keloid scars. Dr. Lin Chen is a Research Associate Professor in the Center for Wound Repair and Regeneration at the University of Illinois at Chicago College of Dentistry, whose research focuses on mucosal and cutaneous wound healing, angiogenesis, and immune response in wound repair.