Phenotype and function of fibroblasts in normally healing wounds

The fibroblast, the major type of cell in granulation tissue and in EGT, changes its phenotype during healing. The phenotype and function of fibroblasts are closely related. When healing is uncomplicated, phenotypes reflect the various needs of the wound as healing progresses, and phenotypes succeed one another as the wound matures. Initially, fibroblasts have a migratory phenotype, allowing them to move from the surrounding tissues into the wound. Migration depends on chemoattractive agents released by platelets and macrophages at the wound’s border in response to injury.⁵ Once at its ultimate destination, the fibroblast changes its phenotype into a proliferative and synthesizing form. Consequently, the number of fibroblasts increases, and ECM is produced. The synthesis of ECM is stimulated by several cytokines, such as transforming growth factor beta (TGF‐β), released either by inflammatory cells or by the fibroblasts themselves.^6,7 Thereafter, fibroblasts can differentiate into myofibroblasts, a phenotype that contains smooth muscle actin filaments and can pull the margin of the wound centripetally via the contractile force exerted by these filaments.⁸ Differentiation of fibroblasts into myofibroblasts and the generation of contractile forces are also stimulated by TGF‐β, but both are inhibited by many other mediators produced in a chronically inflamed environment. After contraction ceases, fibroblasts and myofibroblasts disappear from the wound by apoptosis (i.e. cellular fragmentation followed by phagocytosis by macrophages and activated fibroblasts), and the cellularity of the repair tissue diminishes.^9,10 Apoptosis, a form of non‐inflammatory programmed cellular death, is thus critical to the transition from one phase of repair (proliferation and contraction) to the next (remodeling). Myofibroblasts may represent the terminal differentiation state of fibroblasts, after which apoptosis can occur.

Phenotype and function of fibroblasts in wounds with exuberant granulation tissue

The development of EGT in a wound coincides with a disordered succession of fibroblastic phenotypes. Specifically, the proliferative and synthesizing phenotypes predominate in EGT, while differentiation into contractile myofibroblasts is delayed.³ This is consistent with the microscopic observation in limb wounds that persistent mitotic activity accompanies chaotically arranged myofibroblasts, a pattern not conducive to contraction,^2,3 as well as with the clinical observation that the presence of EGT is often coupled with poor contraction. Confirmation of this phenomenon was established in vitro, where the rate of proliferation of fibroblasts appeared inversely proportional to the capacity of fibroblasts for contraction because rapidly proliferating fibroblasts produced lower contractile forces (Figure 15.3).¹¹ Indeed, because myofibroblasts represent the terminal state of differentiation of fibroblasts, tardy progression to the contractile phenotype also implies that apoptosis will be impaired and high cellularity will be maintained, thereby favoring the development of EGT.

Because EGT occurs primarily in poorly contracting limb wounds, as opposed to efficiently contracting body wounds, investigators hypothesized that fibroblasts from diverse anatomical origins might possess different inherent characteristics accountable for the variable succession of phenotypes.^11,12 It was found, however, that the elevated mitotic activity in limb wounds is not based on innate differences in characteristics of growth between fibroblasts from limbs and those from the trunk because those of limb origin grow significantly more slowly than those of trunk origin when cultured in vitro.^11,12 Additionally, the inadequate contraction seen in limb wounds is not based on a weak inherent contractile capacity of fibroblasts from limbs; fibroblasts from limbs contract more than those from the trunk when cultured in vitro.¹¹

In conclusion, the different phenotypes and functions attributed to fibroblasts from the limb or trunk, as well as the contrasting modes of repair characterizing limb and body wounds, are not based on distinct intrinsic cellular characteristics but, instead, must be the result of other factors. The extracellular environment’s biochemical, molecular, and physical components govern the phenotype of the fibroblast, whereas the phenotype determines the fibroblast’s response to environmental signals.⁸

Inflammatory response

An inefficient inflammatory response to injury may influence the phenotype and function of fibroblasts and thereby play an important role in the development of EGT in limb wounds of horses (Figure 15.3). After trauma, the acute inflammatory response in limb wounds of horses is weaker during the first 3 weeks than that of limb wounds of ponies, and the concentration of TGF‐β in limb wounds of horses is lower during the first 10 days than that in limb wounds of ponies.^3,13,14 TGF‐β1 not only stimulates production of ECM but also favors the differentiation of fibroblasts into myofibroblasts, thereby encouraging wound contraction. An inferior initial concentration of TGF‐β may delay this differentiation, resulting in the presence of fewer myofibroblasts in favor of the rapidly proliferating and synthesizing fibroblast phenotypes. A reduced number of myofibroblasts means that contraction is delayed and inefficient, whereas proliferation of fibroblasts and synthesis of ECM continue unabated (Figure 15.3).

The weak acute inflammatory response seen in wounds of horses was shown to be followed by a persistent or chronic inflammatory response,³ due in part to the continued presence of contaminants and non‐viable tissue not resolved by the initial, feeble inflammatory response. Additionally, a delay in contraction means that the surface area of an open wound remains larger, thus perpetuating the inflammatory response because leukocytes disappear only after epithelium covers the surface of the wound. The substantial presence of leukocytes in exposed granulation tissue may explain up‐regulated synthesis of cytokines in the absence of epithelium¹⁵ and may lower oxygen tension in the wound as a result of the high oxygen consumption by these cells. Persistence of mediators, such as TGF‐β, platelet‐derived growth factor (PDGF), and fibroblast growth factor (FGF), induces fibrosis, whereas prostaglandin (PG)E₁, PGE₂, and interferon (IFN)γ inhibit contraction, while yet others, such as tumor necrosis factor (TNF)α, interleukin (IL)‐1, and IL‐6 do both (Figure 15.4).^16,17 Low oxygen tension additionally stimulates proliferation of fibroblasts and production of ECM.^18,19 The inflammation persisting in limb wounds of horses, therefore, likely enhances the formation of EGT and inhibits contraction, phenomena that are often seen simultaneously in the clinical setting. This hypothesis is substantiated by a recent study confirming that exacerbated and prolonged inflammation in healing wounds on the ears of rabbits favors the development of hypertrophic scarring.²⁰ Moreover, the observation that corticosteroids, potent anti‐inflammatory drugs, control the formation of EGT in the wounds of horses adds further credence to this scenario.²¹

In summary, the combination of an inefficient, weak, acute inflammatory response and the ensuing chronic inflammation in limb wounds of horses delays the differentiation of fibroblasts into myofibroblasts, reducing wound contraction and favoring proliferation of fibroblasts and synthesis of proteins. This leads to a rapid increase in tissue volume by cellular proliferation, rather than a decrease in tissue volume by contraction (Figure 15.5). The chronic inflammation inherent to second‐intention healing in limb wounds of horses, while often unrecognized clinically because of the mild accompanying signs, is no doubt a very important trigger for formation of EGT. The interaction between inflammation, subsequent formation of EGT, and lack of contraction establishes a vicious cycle because these physiologic phenomena stimulate one another.

Local cytokine profile

The aforementioned development of chronic inflammation in limb wounds of horses substantiates several studies documenting a fibrogenic‐rich, local cytokine profile in limb wounds.^13,22–25 One of these cytokines, TGF‐β1, stimulates migration and proliferation of fibroblasts and their production of ECM proteins, such as fibronectin and collagen,⁶ while inhibiting the degradation of ECM.^25,26 It is thus noteworthy that the expression of TGF‐β1 persists in limb wounds throughout the proliferative phase of repair, whereas it quickly returns to baseline values in body wounds after the initial inflammatory phase of healing.^13,22 Persistent production of TGF‐β1 in limb wounds may partially be the work of the fibroblasts within the wound that also express more TGF‐β receptors;^27,28 the signaling components are thus in place to stimulate cellular proliferation and encourage accumulation of components of ECM.

The persistent expression of TGF‐β1 can be explained, at least in part, by various characteristics of limb wounds in horses, such as absence of epithelium, presence of tightly fixed surrounding skin, as well as hypoxia of local tissue. Indeed, the synthesis of fibrogenic cytokines is up‐regulated in the absence of an epithelial cover.¹⁵ Persistent mechanical tension in a wound also plays a role, because mechanical unloading of fibroblasts is required to desensitize cytokine receptors, abrogating cytokine responsiveness and thereby favoring apoptosis.^29,30 Consequently, mechanical stress must be relieved for a wound to progress from granulation tissue to scar, by apoptosis.²⁹ Additionally, the secretion of TGF‐β1 by fibroblasts is strongly stimulated by low oxygen tension.^19,31

Collagen synthesis, deposition, and lysis

Investigators have assumed for some time that aberrant metabolism of collagen plays a key role in the formation of EGT. Theoretically, either abundant synthesis or impaired lysis may lead to excessive accumulation of collagen. The horse forms collagen speedily in response to wounding, indicating a prompt and excessive connective tissue response compared to the response of other species.³² As described earlier, protracted expression of TGF‐β1 in limb wounds may give rise to excessive formation of collagen and other proteins of the ECM.^13,23 Moreover, superior concentrations of type‐I collagen and tissue inhibitors of metalloproteinase (TIMP)‐1 mRNA have been measured in limb wounds compared to body wounds of horses at 1 and 4 weeks of healing.²⁴ Because TIMP‐1 inhibits lysis of collagen, high concentrations present 4 weeks post wounding might favor accumulation of ECM. An imbalance between synthesis and degradation of collagen is likely correlated to the development of EGT, indicating that agents inhibiting collagen synthesis or stimulating the activity of MMPs, in particular collagenase, may represent therapeutic options. Indeed, the Food and Drug Administration approved intralesional collagenase for the treatment of Dupuytren’s disease, a proliferative connective tissue disorder in humans;³³ moreover, this approach is currently under investigation for the management of keloid scarring in humans, a dermal fibroproliferative condition that resembles equine EGT.³⁴ Because an imbalance between synthesis and degradation of a single component of the ECM is unlikely to be the sole basis of formation of EGT, however, the effect of influencing only the metabolism of collagen might be limited.

Angiogenesis and wound oxygenation

Because low oxygen concentrations have been shown to stimulate the proliferation of fibroblasts and their synthesis of components of the ECM, it was postulated that local hypoxia within the granulation tissue of limb wounds of horses might contribute to the development of EGT. This hypothesis was verified via a series of experiments, the first of which mapped the expression of several genes and their corresponding proteins, selected from a larger pool^35–41 because of their known contribution to angiogenesis during wound healing.^42,43 Healing limb wounds of horses were found to be deficient in anti‐angiogenic molecules, compared to body wounds, suggesting that the “control switch” to limit angiogenesis is defective in wounds on the limbs of horses. This coincides with the exacerbated angiogenesis observed clinically in limb wounds,⁴⁴ an important feature of EGT. Although the granulation tissue of limb wounds of horses is characterized by marked vascular regeneration, the lumens of these new microvessels are occluded significantly more than those of microvessels found in body wounds,⁴⁴ owing to hypertrophy of the lining endothelial cells.⁴⁵ This hypertrophy is also observed in keloid scars of humans,⁴⁶ which share numerous clinical and histopathologic features with equine EGT.^47,48

Given this occlusion, the function of new blood vessels in healing wounds of horses was assessed by monitoring derivatives of cutaneous blood flow, namely the temperature of skin and wound⁴⁹ as well as transcutaneous oxygen saturation levels.⁵⁰ Cutaneous wound temperature, and by extension blood flow, was found to be significantly inferior in limb wounds compared to body wounds and even lower in limb wounds predisposed to the formation of EGT.⁴⁹ This data was corroborated using laser Doppler flowmetry.⁵¹ Concomitantly, the degree of oxygen saturation in limb wounds of horses was found to be significantly inferior to that of body wounds during the early period of healing, indicating a temporary, relative state of hypoxia during the inflammatory phase of repair.⁵⁰ Likewise, metabolic disturbances were found, via microdialysis, confirming an inadequate supply of oxygen during healing of equine limb wounds that developed EGT.⁵¹

Because oxygen is required for bactericidal efficiency of leukocytes,^52,53 the relative hypoxia present acutely in limb wounds of horses may explain the feeble yet prolonged inflammatory response in these wounds thought to contribute to the development of EGT.³ Moreover, hypoxia itself has been shown to up‐regulate angiogenic and fibrogenic mediators,⁵⁴ thereby encouraging angiogenesis as well as the proliferation of fibroblasts and their synthesis of components of ECM, substantiating the relationship between the (low) oxygen saturation levels in a wound and the wound’s propensity to become fibro‐proliferative. These findings were confirmed in a study that showed that subjecting cultured equine dermal fibroblasts to hypoxia stimulated the fibroblasts to proliferate and to synthesize ECM while decreasing the turnover of ECM.¹⁹ Interestingly, no behavioral differences in response to hypoxia were observed between fibroblasts originating from the body or the limb, implying that development of EGT does not depend on intrinsic properties of limb fibroblasts but rather on the local environment of the wound.

In view of these findings, one would assume that delivery of supplemental oxygen during the acute inflammatory phase of healing might protect against the development of EGT in limb wounds of horses. Hyberbaric oxygen therapy (HBOT) involves the inhalation of 100% oxygen within a pressurized chamber, which leads to increased tissue oxygen tensions (for more information on HBOT, the reader is referred to Chapter 19).⁵⁵ Topical oxygen therapy (TOT) involves increasing the supply of oxygen directly to the wound by placing an oxygen‐filled bag, boot or extremity chamber around the limb or affected area. HBOT is used to manage wounds in humans but as yet, evidence of its efficacy, based on high‐quality research, is lacking.⁵⁶ Preliminary evidence suggests that HBOT is not indicated for use in horses after full‐thickness skin grafting of uncompromised, fresh, granulating wounds,⁵⁷ but this modality has not been evaluated in the management of other types of wounds in horses. Likewise, a preliminary study on the use of TOT in the management of experimental dermal wounds on the limbs of healthy horses also showed little effect on healing.⁵⁸

Apoptosis

The elevated and persistent mitotic activity characterizing the granulation tissue of limb wounds may relate to deficient apoptosis. This assumption has been partially substantiated by the observation that the balance of apoptotic signals is skewed against apoptosis in limb compared to body wounds of horses.⁴⁴ High concentrations of TGF‐β1 persisting in limb wounds²³ may play a role because this particular cytokine appears to have an anti‐apoptotic effect on fibroblasts.^59,60 Furthermore, in addition to encouraging the synthesis of fibrogenic cytokines, the absence of an epithelial cover, as is characteristic of EGT, debilitates apoptosis because many signals that favor elimination of fibroblasts by this process are normally released by the keratinocyte.⁶¹ Signals that bring about apoptosis also participate in decreasing deposition of collagen, not only by reducing numbers of fibroblasts but also by activating collagenase.⁶¹

If the signal to down‐regulate the activity of fibroblasts and myofibroblasts is delayed beyond a specific point in time, then apoptosis is permanently impaired.⁶¹ This may explain why wounds chronically affected with EGT are unlikely to resolve spontaneously; indeed, the impairment of apoptosis limits the elimination of unwanted cells and subsequent transition to the next phase of repair.

Location of the wound

The likelihood that a wound will develop EGT depends on its anatomic location. Wounds of the body heal quickly and without the formation of EGT in contrast to limb wounds, which heal slowly and are prone to excessive fibroplasia. The exact location of the wound on the limb further influences healing and formation of EGT; wounds over the dorsal surface of the metacarpo/metatarsophalangeal joint (fetlock) heal more slowly than similar wounds located over the dorsal surface of the metacarpus/metatarsus.^62,63 Additionally, limb wounds located on the extensor and flexor surfaces of joints and the heel bulbs appear prone to the development of EGT. These clinical differences are assumed to relate to movement that tears the granulation tissue, inciting more inflammation and cellular proliferation (Figure 15.6). The movement of partially lacerated or frayed tendons may exert a similar influence. Indeed, restricting movement by applying a cast or splint can prevent repeated damage and reduce the formation of EGT in these wounds, despite the fact that the local environment created by the cast would normally favor the formation of EGT.⁶⁴

Another factor enhancing the formation of EGT in the distal aspect of the limb is the relative lack of tissue covering the underlying bone and, subsequently, a reduced vascular bed and relatively poor collateral circulation. Impairment of circulation from trauma results in lower oxygen tension in the healing wound, as addressed earlier, with the ensuing effects on proliferation of fibroblasts and synthesis of ECM.^31,54 Conversely, thick and well‐vascularized musculature covers most structures of the trunk so that perfusion of a wound in this location is not usually substantially disturbed.

Breed

Formation of EGT is influenced by the breed, but not in an exclusive manner; both horses and ponies can develop EGT, although to differing extents.² Ponies form EGT with lesser frequency and quantity, and EGT of ponies tends to disappear spontaneously after the wound is left uncovered. In contrast, horses consistently form more EGT that does not disappear when left uncovered and must be excised. This variable breed predilection might be the confounding factor that explains the conflicting conclusions of past investigations of whether ponies heal with or without EGT.^21,62,64

Factors related to inflammation and infection

Many factors in the wound can stimulate the overproduction of granulation tissue; most of them are related to inflammation and/or infection. A generalized wound infection may not lead to the formation of EGT but might arrest healing, resulting in an indolent wound. On the other hand, local infection related to the presence of bony sequestra, necrotic segments of tendons, ligaments or other tissue, and/or foreign bodies, triggers a chronic inflammatory response leading to the aforementioned cycle, especially when the condition is long‐standing. Similarly, leukocytes are strongly attracted to wounds contaminated with dirt or bacteria, leading to chronic inflammation of the wound.

This emphasizes the importance of thoroughly examining a wound with EGT to identify or exclude factors related to inflammation and infection. If no specific causal factor is incriminated, the inherent chronic inflammatory response commonly occurring in limb wounds of horses may well be the trigger for the formation of EGT. This may, in fact, be the most common cause of EGT and often goes unrecognized because the signs of chronic inflammation may be limited to irregularity of the surface of the granulation bed accompanied by the presence of purulent exudate. These clinical signs, as well as culturing bacteria from the wound, do not indicate that a wound is infected. The wound usually improves dramatically, however, in response to reduction of the bioburden at the surface of the wound, in combination with a single local application of a corticosteroid, a fact substantiating the involvement of chronic inflammation in formation of EGT.

Bandages and casts

Bandages used to cover wounds on the distal aspect of the limb are usually comprised of a primary layer (the dressing), a secondary layer that provides protection and support and can absorb excess exudate, and a tertiary layer that compresses and supports the other layers.⁶⁵ Full‐thickness wounds on the distal aspect of the limb of horses have been shown to be more likely to develop EGT when covered with a bandage or cast than are similar wounds left unbandaged.^{21,27,63–66} This effect is thought to result because bandaging or casting induce the following phenomena: (1) an increase in the oxygen gradient between deeper tissues and the surface of the wound, stimulating angiogenesis;⁶⁷ (2) a reduction in oxygen tension in the wound tissues, enhancing proliferation of fibroblasts;¹⁸ and (3) the creation of a moist, warm, and acidic environment, favoring cellular migration and proliferation. Furthermore, some dressings can irritate the wound, evoking more inflammation, and can cause exudate to accumulate at the wound’s surface. All these features prompt the formation of EGT.⁶² In this respect, management practices can explain, at least partially, why EGT is more often seen in limb wounds because these wounds are usually bandaged whereas body wounds are not. Interestingly, body wounds, when bandaged, can also form EGT, albeit to a lesser degree. EGT would likely form more often in body wounds, if these wounds were routinely bandaged. In conclusion, management practices contribute to the effect of location on the formation of EGT.

The precise effect of bandages on the development of EGT is especially dependent on the type of dressing used. In general, the more occlusive the dressing, the higher is the incidence of EGT. Synthetic, occlusive dressings were shown to significantly stimulate the formation of EGT, thereby increasing the need to trim granulation tissue, compared to other less occlusive dressings.⁶⁸ This effect may be explained by an excessive accumulation of exudate beneath the occlusive dressing, which in turn, favors bacterial proliferation and encourages inflammation. Some dressings that are not occlusive by nature can nevertheless “seal” the wound because they insufficiently absorb exudate, causing it to accumulates at the wound’s surface, thereby creating a barrier to diffusion. Likewise, the surface of wounds dressed with a non‐occlusive dressing, such as gauze, may become occluded when exudate passing through the gauze into the secondary layer of the bandage causes the latter to become occlusive. Several topical wound‐care products applied to the surface of a wound, particularly those of a fatty nature, may also cause occlusion.

Conversely, semi‐occlusive dressings with a high absorptive capacity, such as foams, provide an environment for the wound that is less likely to induce the formation of EGT because exudate and bacteria are wicked away from the wound (the reader is referred to Chapter 6 for more information on wound dressings).

A notable exception to the general rule, “the more occlusive the dressing, the higher the incidence of EGT,” pertains to an occlusive silicone gel sheet dressing (CicaCare®, Smith & Nephew) that was better able to prevent the formation of EGT in limb wounds of horses than was a permeable dressing (Melolite®, Smith & Nephew).⁶⁹ Silicone gel sheets are currently the gold standard dressing for managing excessive scarring of the skin in humans.^70,71 Although the mode of action of the silicone membrane is not entirely understood, the dressing has been shown, in people suffering from hypertrophic and keloid scars, to increase tissue temperature⁷² and hydration,⁷³ thereby enhancing the activity of collagenase, which favors remodeling of ECM.⁷⁴

Although bandages and casts can promote the formation of EGT, this does not preclude their use in managing wounds. To the contrary, bandages exert many positive influences on healing. They keep the wound clean and prevent contamination and irritation by environmental factors, such as dirt and straw, which induce inflammation. They facilitate administration of local wound therapies, stimulate more rapid formation of granulation tissue, which is initially required for healing, accelerate epithelialization by creating a moist environment, encourage the development of a more cosmetic scar,⁶⁴ reduce the risk for sarcoid transformation of the wound, and protect the wound from additional trauma. Additionally, casts restrict movement in highly mobile regions, thus reducing disruption of the healing process. All these functions limit the development of EGT. Bandages and casts are, therefore, an important component in the management of wounds healing by second intention, because the aim to support the overall healing process largely surpasses the need to prevent EGT.

Iatrogenic factors

The way in which a wound is managed in the early stages has a dramatic effect on the time required for healing, as well as on the formation of EGT. Application of non‐physiologic materials, including powders and chemicals (e.g., antiseptics), as well as some antibiotics, may adversely affect wound healing (Figure 15.7). Additionally, application of caustic substances, including copper sulfate, nitric acid, acetic/malic acid mixtures, silver nitrate, triple dye, supersaturated potassium permanganate, sodium hypochlorite (Dakin’s solution), lye, and many other home remedies, in an effort to prevent or treat EGT, seriously delays repair. These caustic agents induce necrosis not only of the granulation tissue but also of the migrating and proliferating epithelium.^62,75 Cryogenic surgery applied to granulating wounds similarly delays healing.⁶⁴ The resultant necrosis encourages a chronic inflammatory response and the release of many mediators that inhibit wound contraction and overstimulate cellular proliferation, as previously outlined. Healing by epithelialization and contraction is arrested, and the stimulus for formation of granulation tissue accrued, leading to a recurrence of EGT. Cautery of any type, therefore, delays healing and promotes more EGT. Although wounds treated in this way eventually heal, they are frequently characterized by unacceptable scarring.⁷⁵ The aforementioned problem is unfortunately still encountered in some equine practices, because many owners attempt to manage a wound themselves prior to consulting a veterinarian.