Scar formation is a major clinical problem resulting in adverse cosmesis, loss of function, especially if over joints, and hindrance of growth in children. Scars also have a dramatic impact on the patient’s quality of life, and have been associated with anxiety, social avoidance, and depression.1 Therefore, prevention of scar formation has long been important. Scar reduction is not only important to dermatologists, but is also salient in many other conditions, such as adhesions and strictures resulting from surgical procedures in abdominal and pelvic cavities, spinal cord ruptures, scarification of hand tendons after injury, corneal abrasions, glomerulonephritis, cirrhosis, human vascular restenosis lesions, myocardial infarction, systemic sclerosis, and diffuse fasciitis. The principles of anti-scarring therapy based on modulation of pro-scarring vs. anti-scarring factors appear to be promising for all the above-mentioned conditions, indicating that studies in the skin could have a broader clinical application.
Scar and fibrosis are the end result of surgical and nonsurgical skin injury. Aggressive wound healing may have once offered an evolutionary advantage for survival at the expense of scar formation. Amazingly, fetal cutaneous wounds, especially in the first 6 months of gestation, heal without scar formation.2 Many clinicians hope that understanding the remarkable reparative capabilities of the fetus may lead to the development of new wound-healing therapies that reduce or prevent scar formation and fibrosis.
Fetal monkey lip incisional wounds heal with restoration of normal appendage and dermal collagen architecture in midgestation. At the start of the third trimester, these wounds do not restore appendage (hair follicle and sebaceous gland) architecture, but still heal with a normal collagen pattern. Thus, a “transition wound” phenotype occurs. By the mid-third trimester, the wounds heal with a typical scar pattern, i.e. no appendages and collagen scar.3 The transition point in a human fetal skin model also occurs after the second trimester of gestation.4
It is notable that not all fetal tissues share the anti-scarring properties of fetal skin; fetal wounds in the diaphragm and the gastrointestinal tract heal by fibrosis and contraction, as in the adults.5,6 Fetal cutaneous wound healing is not only scarless but also rapid. The rapid epithelialization of fetal wounds may occur in part because of early deposition of tenascin and fibronectin, which are thought to be necessary for migration and cell anchoring, respectively.7 As discussed below, overexpression of vascular endothelial growth factor (VEGF) may be another reason for rapid healing of fetal wounds. How is scarless wound healing different than the scarring wound healing? In scarless wounds, collagen is rapidly deposited in a fine reticular pattern indistinguishable from uninjured skin. In contrast, adult scarring wounds have disorganized and thick collagen bundles with more collagen cross-linking.7
Interestingly, amniotic fluid is neither essential nor sufficient for scarless repair. Fetal marsupials develop outside the uterus in a maternal pouch and heal cutaneous wounds without scar.8 Adult sheep skin transplanted onto the backs of fetal sheep bathed in the amniotic fluid of the intrauterine environment heal with scarring of incisional wounds.9
Scarless wound-healing physiology
Transforming growth factor-β (TGF-β)
Transforming growth factor-β is secreted by most cells involved in wound healing, including neutrophils, lymphocytes, macrophages, keratinocytes, and fibroblasts. Transforming growth factor-β is first released from degranulating platelets.14 Interestingly, TGF-β upregulates its own production in an autocrine pattern, leading to TGF-β overproduction and scar formation.14 Transforming growth factor-β is a potent chemoattractant of macrophages, neutrophils and fibroblasts, and stimulates extracellular matrix synthesis and prevents its degradation by upregulating the expression of tissue inhibitors of metalloproteinases (TIMPs) and downregulating the expression of proteases. Three highly homologous TGF-β genes in mammals, designated TGF-β1, -β2 and -β3, have been identified.14 Shah et al.15 treated adult rodent wounds with isoform-specific neutralizing antibodies either alone or in combination. Exogenous addition of neutralizing antibody to TGF-β1 alone resulted in some reduction in the inflammatory and angiogenic responses, as well as reduction of extracellular matrix deposition in the early stages with a marginal reduction in cutaneous scarring. By contrast, neutralizing antibody to TGF-β2 alone had little effect on the inflammatory or angiogenic responses, and no effect on the resultant scar. However, when neutralizing antibodies to TGF-β1 and TGF-β2 were administered together, the synergistic effect resulted in a dramatic amelioration of scar formation. Surprisingly, exogenous addition of TGF-β3 to cutaneous wounds in adult rodents also produced the same effects. Wounds treated with TGF-β3 showed a marked reduction in the immunoreactivity for TGF-β1 and TGF-β2 on Days 5 and 7 postwounding compared with untreated control wounds. TGF-β3 is thus believed to downregulate TGF-β1 and TGF-β2.14
Fetal wounds have higher ratios of TGF-β3 to TGF-β1 and TGF-β2, favoring less scar formation.16,17
Additionally, in a recent study by Rolfe et al.18 it has been shown that fetal fibroblasts, in comparison with mature fibroblasts, respond differently to TGF-β1 stimulation and fail to demonstrate the TGF-β1-induced production of collagen (mRNA and protein). This different response is due to a comparatively short-lived or rapid phosphorylation of several components of the TGF-β1 signaling pathways.
Interestingly, healing of oral mucosal wounds produces less scarring compared with skin. Expression of Integrin αvβ6 is increased in oral mucosal epithelium during wound healing; this protein enhances the secretion of both TGF-β1 and TGF-β3. In a recent study by Eslami et al., expression of TGF-β molecules and β6 integrin was assessed in experimental human gingival wounds, and the gingiva and skin of red Duroc pigs by means of real-time polymerase chain reaction, gene microarrays, and immunostaining.17 As with human wounds, β6 integrin expression was upregulated in the pig wounds 7 days postwounding, and remained enhanced for more than 49 days. The expression levels of β6 integrin and TGF-β3 were markedly higher in the pig gingival wounds compared with cutaneous wounds. It was concluded that the prolonged expression of Integrin αvβ6 resulting in a higher concentration of activated TGF-β3 protected gingival wound epithelium from scar formation.
Decorin, fibromodulin, and lysyl oxidase
Decorin is a proteoglycan component of connective tissue. It binds to type I collagen fibrils and plays a role in matrix assembly. Decorin’s name is a derivative of the fact that it “decorates” collagen. Fibromodulin is a small interstitial proteoglycan, which may participate in the assembly of the extracellular matrix as it interacts with type I and type II collagen fibrils and inhibits fibrillogenesis in vitro. It may also downregulate TGF-β activities by sequestering TGF-β in the extracellular matrix. Decorin is downregulated in fetal wounds, while fibromodulin is upregulated.2
Lysyl oxidase levels also increase during fetal skin development, perhaps resulting in higher collagen cross-linking.2
Matrix metalloproteinases (MMP) and TIMPs
Scarless wounds have a higher ratio of MMP to TIMP, favoring remodeling and less accumulation of collagen.2,19 This may be due to decreased expression of TGF-β1, as it decreases MMP and increases TIMP expression, favoring collagen accumulation and scarring.20
Platelet-derived growth factor (PDGF) and fibroblast growth factor (FGF)
Scarring wounds have higher levels of the profibrotic cytokines PDGF and FGF. Platelet-derived growth factor is a potent mitogen and chemoattractant for fibroblasts.2
Hypoxia and VEGF
The fetal skin is relatively hypoxic. Midgestational fetal lamb tissue pO2 is about 16 mmHg, whereas adult tissue pO2 ranges from 45 to 60 mmHg. TGF-β1 production by fetal fibroblasts may be blunted in hypoxemic conditions.14 Physiologically low oxygen concentrations in fetal skin upregulate hypoxia-inducible factor 1, which is a potent transcriptional regulator of oxygen-dependent genes such as VEGF and TGF-β3.21 Vascular endothelial growth factor, a mitogen for endothelial cells, increases twofold in scarless wounds, while its expression remains unchanged in scarring fetal wounds. Interestingly, one novel, non-protein angiogenic factor is nicotine, which acts through nicotinic acetylcholine receptors to stimulate angiogenesis. The use of nicotine accelerates angiogenesis and wound healing in the diabetic mouse model.20 On the other hand, it is shown that endostatin, an angiogenesis inhibitor, minimizes scarring of mouse wounds.20 It is difficult to interpret these results given that VEGF is expressed more in scarless healing. Therefore, the role of VEGF in scarless healing needs more exploration.
Fibroblasts and myofibroblasts
Fibroblasts are major players in wound healing after developmental maturity; by gaining contractile proteins such as α-smooth muscle actin they transform into myofibroblasts. These cells have certain clinical features of smooth muscle cell, and contribute to scar formation and contraction; however, their presence during fetal wound healing is controversial. Some studies have shown their absence22 while others, such as Cass et al.,23 have shown their presence. In the above-mentioned study by Rolfe et al.,18 fetal fibroblasts exposed to TGF-β1 have been found to be capable of maturation into myofibroblasts; however, this differentiation was found to be quicker and more transitory compared with mature fibroblasts with less production of collagen.
The contractile forces generated by myofibroblasts may alter the orientation of collagen fibrils and contribute to scarring. Adult fibroblasts need to proliferate first before being able to lay down collagen, while fetal fibroblasts are capable of proliferating simultaneously while making collagen. As a result, collagen deposition is delayed in adult wound healing, contributing to scar formation.7,12 Additionally, fetal fibroblasts at early gestation express a lower level of TGF-β receptor II than fetal cells later in gestation. It is also demonstrated that receptor tyrosine kinases express differentially between fetal and adult rat fibroblasts, suggesting that further elucidation of the TGF-β receptor signaling process in fetal fibroblasts may help to understand how fetal cells respond to TGF-β differently from adult cells.24 Moreover, as mentioned earlier, fetal fibroblasts have more HA receptors than adult fibroblasts.
Recently, Kong et al.25 identified a novel, previously unidentified group of E-cadherin-positive cells in the blood of fetal and adult mice, and named them “Dot cells”. Dot cells have a tiny dot shape with a diameter between 1 and 7 μm. It is suggested that Dot cells are relatively primitive cells or stem cells because of their unusually small size and expression of stem cell markers such as E-cadherin, integrin β1 and CD34 by these cells. The percentage of Dot cells in fetal mice blood is more than 20 times higher compared with adult blood. Dot cells migrate to wounds and differentiate into dermal cells, which release less interstitial collagen and reduce scarring. Transplantation of Dot cells to adult mice heals skin wounds with less scarring due to reduced smooth muscle actin and collagen expression in the repair tissue. These results infer that Dot cells are a previously unidentified component in scarless wound healing.
Inflammation and HA
Scarless fetal wounds have less inflammatory infiltrate, which may lead to improvement in the final wound-healing process. Decreased inflammation may be partly due to decreased fetal platelet degranulation and aggregation with the resultant lower levels of the chemoattractants TGF-β and PDGF.2 Moreover, fetal neutrophils may not possess the chemotactic ability of adult neutrophils.25 Both interleukin (IL)-6 and IL-8, which are important in chemotaxis and activation of inflammatory cells, are significantly lower in early fetal fibroblasts. Treating adult mouse wounds with IL-10, which has an anti-inflammatory effect through decreased production of IL-6 and IL-8, reduces inflammation and helps produce scarless healing. This may have therapeutic implications in human adult wounds.2,26–28
Insomnia, through increasing IL-6 levels,29 might predispose patients to scar formation. If proven by controlled studies, this concept can have therapeutic implications, especially after myocardial infarction, severe burns and trauma.
High levels of HA may also contribute to decreased inflammation of fetal wounds. In in vitro systems, the chemotactic and random migration of white blood cells can be inhibited by HA. The phagocytic activity of mononuclear phagocytes is also inhibited by even a relatively low concentration of HA (≤0.05 mg/ml). Therefore, HA exerts biomechanical regulation over processes such as inflammation through its ability to modify the activity of cells involved in the inflammatory response.13 The HA molecular network is able to exclude large molecules, such as fibrinogen and other proteins. Steric hindrance by the HA matrix may alter the chemotactic gradient, and thereby influence the magnitude and nature of the inflammatory response. Hyaluronic acid also serves as a high-capacity free radical scavenger. It is suggested that the presence of HA has a physical effect on the fibrin matrix that forms. Hyaluronic acid also limits fibrin formation through binding to the fibrinogen molecule, with the resultant prevention of excessive collagen deposition.13
The balance of cytokines in the fetus favors HA expression. Proinflammatory cytokines, such as IL-1 and tumor necrosis factor-alpha, which downregulate HA expression, are underexpressed in the fetal wounds, whereas in adults there is upregulation of these cytokines in response to wound healing. As mentioned previously, HASA also contributes to the higher production of HA in fetal skin.
Angiotensin-converting enzyme (ACE)
Recent studies indicate that upregulated ACE may participate in cutaneous pathological scar formation,30 suggesting that ACE inhibitors may exert an anti-scar effect. Moreover, ACE inhibitors also retard Smad3 function, which plays an important role in the TGF-β-induced fibrosis.31
Homeobox genes are transcription factors that are implicated in the patterning and cell type specification events during development. Human homeobox genes MSX-1, MSX-2, and MOX-1 are differentially expressed in skin development. Moreover, fetal scarless repair is associated with decreased expression of HOXB13 and increased PRX-2 expression.2 This implies that PRX-2 activation is an important stimulant to dermal generation. Conversely, HOX-B13, which is strongly expressed in normal second trimester fetal skin, is markedly downregulated in response to wounding. Thus, HOX-B13 may be an inhibitor of dermal proliferation, and its constant expression may be involved in maintaining a static dermal architecture rather than promoting dermal growth.32
Hox genes, another subgroup of homeobox genes, encode for a family of transcription factors that are major regulators of tissue migration and cell differentiation during embryogenesis. In a recent study, Jain et al.33 reported the increased expression of Hox gene Hoxd8 in excisional wounds of mid-gestational mice compared with the normal skin samples of mid-gestational mice and excisional wounds of late-gestational mice; therefore, it is proposed that this gene may have anti-scarring effects in fetal wound healing. Conversely, expression of another Hox gene Hoxd3 was found to be increased in both excisional wounds and normal skin samples of mid-gestational mice compared with late-gestational mice; hence, this gene is considered to be expressed constitutively in the skin of mid-gestational mice.
It is likely that fetal mammals have the ability to heal large skin defects made early in gestation because transcription factor patterning genes, such as homeobox genes, are more active in the fetal environment. As a result, researchers have hypothesized that transcription factors, like homeobox genes, may be the “first domino” in the fetal cutaneous wound repair regulatory cascade.32
How these homeobox genes may coordinate scarless fetal wound repair is currently under investigation. Several possible targets have been identified, including the promoter regions of members of the TGF-β superfamily, various cellular adhesion molecules, and cell surface proteins such as integrins.32
Improvement of scar cosmesis with age
Wound healing is delayed in old age, but there is less scarring. Diminished inflammatory responses due to impaired macrophage and T-cell functions, loss of fibroblast responsiveness and motility, as well as reduced quantity and distribution pattern of various growth factors, including TGF-β, and their receptors may contribute to this effect. Estrogen upregulates TGF-β1 through binding to its receptors on fibroblasts. The decline of estrogen after menopause may explain decreased scarring in the postmenopausal period. Systemic hormone replacement therapy for 3 months accelerates re-epithelialization and collagen deposition of acute wounds in postmenopausal females.34
In normal aging skin, the levels of proteases such as MMPs increase, while the levels of proteolytic inhibitors (e.g. TIMPs) decrease, leading to less scar formation.35,36
TGF-β3 increases at Day 7 postwounding in old mice, while TGF-β1 and TGF-β2 increase at all time points in the wounds of younger mice. This may also explain the better cosmesis of wound healing in aging.12
The differences between fetal and adult wound healing discussed above are summarized in Table 1.
Table 1. Comparison of fetal vs. adult wound healing
|Wound healing||Scarless, rapid||Scarring, slow|
|Collagen quality||Fine, reticular, less cross-linking||Thick, disorganized, more cross-linking|
|Collagen type III to type I ratio||Higher||Lower|
|Tenascin and fibronectin deposition||Early||Delayed|
|Lysyl oxidase levels||Lower||Higher|
|TGF-β1 and -β2 to TGF-β3 ratio||Lower||Higher|
|TGF-β1-induced collagen production||Absent||Present|
|Decorin and fibromodulin levels||Lower/higher||Higher/lower|
|MMP to TIMP ratio||Higher||Lower|
|PDGF and FGF levels||Lower/lower||Higher/higher|
|Fibroblast HA and TGF receptors||More/less||Less/more|
|HA (and tissue fluidity)||More||Less|
|Inflammation (and inflammatory cytokines)||Inconspicuous||Conspicuous|
|Homeobox genes||More active||Less active|
Accepted, potentially useful and questionable strategies for minimizing scars
• The use of proper surgical techniques, which ensure minimal tension and inflammation, and avoiding incisions over the midchest and joints, which are scar-prone sites, can lead to better cosmetic results.34
As mucosal wounds heal with reduced scar formation when compared with skin,26 when possible there should be consideration of endoscopic surgical technique.37
Larger wounds may be more likely to heal with scar formation. It is not unreasonable to assume that larger wounds may extend the time of the healing response, thus exposing wound tissues to a different extracellular matrix and growth factor profile. In addition, the larger excisional wounds may stimulate the formation of myofibroblasts in the wound, resulting in scar formation.32
• Occlusive therapy can reduce scar formation. Occlusive therapy includes silicone gels and sheets, non-silicone occlusive sheets, cordran tape and Scarguard MD® (Redrock Laboratories, Great Neck, NY, USA; see below). Occlusion can reduce pro-inflammatory/pro-fibrotic cytokine levels. Increased temperature, even 1 °C, due to occlusion can upregulate collagenase expression.34
Topical silicone gel has been employed for scar reduction since the 1980s. It is probably of value by enhancing wound hydration as occlusive therapy with both silicone- and nonsilicone-based therapies appears to be of equal value.38,39
Scarguard MD® is a new topical over-the-counter medication containing silicone, vitamin E and hydrocortisone. In a pilot study on 12 patients, Scarguard MD® was applied twice daily after removal of a mole, and nothing was applied after removal of a second mole. After 2 months, nine out of the 12 patients described a reduction in erythema and better cosmesis of the scar at the site treated with Scarguard MD® compared with the site that was not treated with Scarguard MD®. These results must be interpreted with caution, however, as the improved cosmesis may be the result of occlusion of a wound rather than the Scarguard MD product. An in vitro study showed that Scarguard MD® may stimulate the release of inactive collagenase precursors that may inhibit scar formation and reduce existing scars.34
• Seprafilm (Genzyme Corporation, Cambridge, MA, USA) is a type of absorbable barrier film composed of un-crosslinked, carboxymethylcellulose/sodium hyaluronate.40 After application, by absorbing moisture from the application site, the dry film acquires a gel consistency over 24 hours.40 This product decreases the incidence of adhesion formation, and enhances the repair of damaged tissues by acting as an occlusive barrier between damaged tissue sites.40 Several studies have reported the effectiveness of Seprafilm in reducing postoperative intra-abdominal adhesion formation.40–42 This product has also been shown to facilitate ileostomy closure.43
• Topical application of HA or saponins, which upregulate HA, could exert an anti-scarring effect.44 The effects of topical exogenous HA on the healing of dermal wounds have been investigated by many independent investigators in recent years. Most researchers have found that HA provides a beneficial effect with regard to the quality of tissue repair scar formation. The results indicate that the greatest benefit is achieved using highly purified, high-molecular-weight HA at a concentration more than 1 mg/ml, and under conditions in which the HA is maintained at the wound site on a continuous basis for a prolonged period of time. In studies in which HA failed to promote or enhance the wound repair process, testing conditions were suboptimal. Low-molecular-weight and/or very dilute HA preparations rapidly diffuse from the injured site. The purity of exogenous HA is critical in order to limit contaminating substances, such as proteins, endotoxin, etc., which may be inflammatory in nature and may produce undesirable biological effects.13
The synthesis of HA appears to be limited by glucosamine availability.45 Thus, administration of adequate amounts of glucosamine by mouth during the first few days after surgery or trauma has been hypothesized to enhance HA production in the wound, thereby promoting swifter healing and also diminishing complications related to scarring, such as adhesion bands.45
• For reduction of scar formation, clinical trials for the use of TGF-β3 and neutralizing antibodies to TGF-β1 and TGF-β2 are being conducted.46 Neutralizing antibody-treated wounds contained fewer macrophages, less angiogenesis, and less collagen and fibronectin. Interestingly, the tensile strength of the scars of these wounds was comparable with controls.47 Whether these antibodies are capable of reducing scar formation should be answered by the above-mentioned trials. Activation of latent TGF-β requires binding of the latent TGF-β to the mannose 6-phosphate/insulin-like growth factor-2 receptor in the presence of plasmin/urokinase. Activation could be blocked by mannose 6-phosphate or by antibodies to the mannose 6-phosphate receptor.14
Another potential target of therapy is connexin 43 (Cx43), which mediates TGF-β signaling. Studies on Cx43 antisense-based gene therapy to prevent scar formation are under way. As lithium increases Cx43 expression, patients who are taking it may exhibit increased susceptibility to scar formation. MicroRNAs, which are single-stranded RNA molecules partially complementary to one or more messenger RNA molecules, may be employed to downregulate gene expression.46
Smad3 plays an important role in the TGF-β-induced fibrosis. Downregulation of Smad3 expression in fibroblasts by small interfering RNA (siRNA) can significantly decrease procollagen gene expression and attenuate the process of fibrosis. RNA interference is a process in which brief RNA sequences, called siRNA, block signals from a particular gene. This process, called gene silencing, inhibits the gene from carrying out its function. However, this approach is facing some challenges, such as the need to develop improved siRNA delivery strategies that combine high specificity and efficiency with a low immunostimulatory and tumorigenic potential.46,48
• Tamoxifen is shown to downregulate the production of TGF-β1, basic FGF and VEGF.49,50 Some in vivo reports indicate that it delays wound healing but ameliorates scar formation.46
• Starting at the day of suture removal, 585-nm pulsed dye laser treatment has improved the quality and cosmetic appearance of surgical scars in a controlled study on 11 patients.51 Both 585-nm pulsed dye laser and the 1064-nm long-pulsed neodymium : yttrium-aluminum-garnet laser treatment of acne scars in the same patient showed both lasers to be effective modalities for the treatment of acne scars.52
• Botulinum toxin has shown improvement in several patients when injected into the muscle underlying a wound immediately after surgery.53 A recent evaluation of 19 patients found botulinum toxin type A injections of value in the treatment of hypertrophic scars.54
• An investigator-blinded, controlled study showed that the onion extract gel significantly improved scar softness, redness, texture and global appearance at the excision site.55
• A double-blind, randomized, placebo-controlled study did not demonstrate efficacy of imiquimod, an anti-fibrosis cytokine inducer, on scar cosmesis.56
• Vitamin E, a TGF-β inhibitor, was shown to be ineffective topically.57
• In one report, cyclooxygenase (Cox)-2 inhibition by celecoxib, which can decrease wound inflammation, was shown to decrease scar deposition,58 but work by others suggested that antagonism of Cox-2 had no effect on the morphology of healed wounds.59
• Commercial products made by snail’s secretion, such as BIO SKIN CARE™ Cream (Andes Natural Skin Care, Carson City, NV, USA) and Eilcina are claimed to have anti-scarring properties. In support of this claim, an early study reported that a secretion from the mollusk Cryptomphalus aspersa (SCA) stimulated skin regeneration after wound-healing impairment due to acute radiodermatitis.60 Moreover, in a recent study, Brieva et al. reported that therapeutic properties of SCA in skin regeneration were due to the presence of antioxidant substances in this compound; SCA was also reported to have extracellular matrix remodeling capabilities.61
• Importantly, despite marketing claims, there is no evidence that acceleration of healing with some agents, such as Cicamosa emulsion (Lutsia, Paris, France), Cicalfate cream (Avene, Paris, France), trolamine (Biafine Genmedix, France) topical emulsion, or other commercial repair creams is associated with decreased scar formation.62,63 Even if these agents do speed healing time, there is no known correlation between healing rate and the degree of scar formation. For example, postmenopausal wounds heal slower but with less scar formation in comparison to premenopausal wounds.34,64 It should be noted that promotion of fibroblast activation as some of these products claim may theoretically induce more scarring.