A Review of Epidermal Maturation Arrest: A Unique Entity or Another Description of Persistent Granulation Tissue?

Abstract

Objective: To conduct a review of reported cases of epidermal maturation arrest and to compare their clinical and histological descriptions with that of persistent granulation tissue with a focus on diagnostic methods and response to treatment. Methods: The authors performed a literature search within Pubmed, Embase, Google Scholar, and Web of Science for all reported cases of epidermal maturation arrest under the terms “epidermal maturation arrest,” “epidermal arrest,” “epidermal maturation,” and “re-epithelialization maturation arrest.” They reviewed the clinical and histological presentation of hypergranulation tissue as well as the evidence for the most widely used treatments. Results: There is only one case series and one case report of epidermal maturation arrest, and the former gives the most detailed clinical and histological description including response to treatment. The clinical description, histological findings, and response to treatment of all cases are comparable to that of persistent granulation tissue and there is no histological or cytological data provided to support that epidermal maturation arrest exists as a distinct entity. Conclusion: Among the cases of epidermal maturation arrest reported in the literature, there is insufficient evidence that keratinocytes acquired a state of arrest in their migration. Rather, the described cases appear to have been complicated by persistent granulation tissue, a well-known aberration in wound healing.

The formation of persistent granulation tissue, also referred to as hypergranulation tissue, proud flesh, exuberant granulation, or hypertrophic granulation,1 is a complication of surgical wound healing in which granulation tissue fills beyond the height of the wound defect and prevents keratinocyte migration from wound edges to complete epithelialization. It presents as exophytic, erythematous, and friable tissue with increased vascularity2 and represents an aberrant response of the proliferative phase of wound healing. Histologically, it can resemble a pyogenic granuloma with an increase in fibroblasts and endothelial cells.3 With treatment of the exuberant tissue utilizing chemical cautery with silver nitrate or topical corticosteroids, epithelialization and migration of keratinocytes from the wound edges may proceed to complete the healing process.4,5

In 1999, a new complication of cutaneous surgery known as epidermal maturation arrest (EMA) was described in a case series of four Mohs defects that failed to heal by second intent despite treatment with topical corticosteroids and oral nonsteroidal anti-inflammatory drugs (NSAIDs) (Table 1). In all four cases, the surgical wounds failed to re-epithelialize at subsequent visits that ranged from three to six months after surgery and during which the defects contained only clinically apparent granulation tissue that did not respond to initial treatments. The authors cited histological evidence, specifically a lack of particular cytokeratins otherwise found in normally healing wounds, to conclude that delayed healing may have resulted from either abnormal epidermal migration or defective proliferation of the keratinocytes at the wound edges.

TABLE 1.

Clinical reports of epidermal maturation arrest

PATIENT DESCRIPTION	PROCEDURE, DIAGNOSIS AND LOCATION	INITIAL DEFECT SIZE	FOLLOW-UP #1	TREATMENT #1	FOLLOW-UP #2	TREATMENT #2	FOLLOW-UP #3	TREATMENT #3	FINAL OUTCOME
38-year-old female6	Status-post MOHs on left temple for basal cell carcinoma arising in sebaceous nevus	8.5 x 2.5cm	At 6 weeks post- operatively: Defect: 7.8 x 2.0cm; Granulation tissue appeared glossy; Biopsy showed granulation tissue	Silver nitrate and biological dressings	At 10 weeks post- operatively: No change in granulation tissue	Silver nitrate and multiple biologic dressings	At 6 months post- operatively: Defect: 6.2 x 1.6cm; Repeat biopsy showed granulation tissue	Six weeks of temovate gel to remaining defect	After 6 weeks of temovate gel, the final defect measured 1cm
44-year-old male6	Status-post MOHs on left forehead and scalp for basal cell carcinoma and squamous cell carcinoma	12.5 x 7.4cm	At 3 weeks post- operatively: Defect: 12.2 x 7.4cm; Granulation tissue appeared glossy	Silver nitrate applied to foci of over- granulation	At 8 weeks post- operatively: Defect: 12.0 x 6.7cm; Granulation tissue unchanged	Temovate ointment under occlusion until 10 weeks post- operatively	At 12 weeks post- operatively: Defect: 10 x 5.8cm	Four weeks of ibuprofen 200mg po TID	At 16 weeks post- operatively: Few focal erosions remaining; At 6 months post- operatively: Healing complete
59-year-old male6	Status-post MOHS to right cheek for recurrent basal cell carcinoma	6.5 x 6.0cm	At 3 weeks post- operatively: Defect: 5.2 x 4.0cm	Not specified	At 3 months post- operatively: Defect: 2.5 x 1.7cm; Granulation tissue with glossy appearance	Ibuprofen po BID with nonstick wound dressings	After 2 weeks of ibuprofen po BID, the defect had re-epithelialized	NA	Two weeks of NSAIDs brought about completed re-epithelialization; topical steroids not applied at all
80-year-old Caucasian female6	Donor SPSG site from the thigh to close MOHS on the right cheek	SPSG site on the thigh: 7.8 x 5.4cm	At 3 weeks post-operatively: defect appeared hypergranulated	Silver nitrate	At 6 weeks post-operatively: Small area of mature epidermis on lower half of defect	Silver nitrate and ibuprofen 200mg po BID	At 12 weeks post-operatively: Defect: 2 x 1cm	Ibuprofen discontinued and halobetasol proprionate ointment applied	At 15 weeks post-operatively: Re-epithelialization complete
76-year-old female8	Status-post laser vaporization for Bowen’s disease on the frontal scalp in two parts; each half of the lesion treated and separated by 6 months	Defect after treatment not specified Silver nitrate applied for unspecified time	At 6 months post-operatively: Granulation tissue persisted; Biopsy showed erosion with edema of papillary dermis and mixed perivascular infiltrate	At 9 months post-operatively: Diflorasone ointment prescribed	At 10 months post-operatively: Re-epithelialization was complete	NA	NA	NA	After 1 month of applying diflorasone ointment, epithelialization was complete

This paper reviews all published cases of epidermal maturation arrest found within Pubmed, Embase, and Web of Science under the search terms “epidermal maturation arrest,” “epidermal arrest,” “epidermal maturation,” and “re-epithelialization maturation arrest.” Secondly, it compares all reported cases of epidermal maturation arrest with that of hypergranulation tissue, with a focus on histological description and response to treatment. Finally, the biochemical changes associated with keratinocyte migration, specifically the concept of epithelial to mesenchymal transition, are reviewed as cytokeratin profiling is essential to assess whether keratinocytes in a healing wound have been programmed to undergo migration and complete re-epithelialization.

A LITERATURE SEARCH FOR EPIDERMAL MATURATION ARREST

In 1999, Jaffe et al6 described four clinical cases of surgical wounds left to heal by second intent, which remained non-epithelialized after three to six months of follow-up. The four wounds finally responded to treatments with NSAIDs, silver nitrate, and topical steroids in two cases, silver nitrate and topical steroids in one case, and NSAIDs only in the fourth case (Table 1). In only one case were biological dressings first applied, which did not result in wound healing until a topical corticosteroid was subsequently applied.6 A biopsy was performed in only one of the four cases as well as in another similar case that was also treated by the authors, but not included in the report. The two biopsies were reportedly consistent with granulation tissue although the exact location from where the biopsies were obtained was not specified, nor were the histological findings detailed. Further, the authors cited negative cytokeratin stains of the biopsy specimens as evidence that keratinocytes had failed to migrate. They concluded that EMA occurs in a post-surgical wound that maintains persistent granulation tissue on sun-damaged skin, but fails to completely re-epithelialize despite a lack of clinically apparent infection. They explain that anti-inflammatory treatments, both topical corticosteroids and NSAIDs, were effective because they upregulate repair phase cytokines and block phospholipase A2 in the arachadonic acid cascade, respectively.6

There are only two other reported cases of EMA. Torne et al7 described the development of a papular cutaneous eruption in a patient after treatment with polychemotherapy for systemic malignant histiocytosis. In the abstract, the authors stated that the biopsy from these lesions was consistent with EMA, but then further described the histological findings as containing “atypical keratinocytes, mitotic figures, and the presence of “starburst” cells, associated with an interface vacuolar dermatitis.”7 The histological and clinical descriptions of this single case do not appear to be relevant to a non-healing wound.

Mandrea et al8 reported a case of a scalp wound that failed to heal six months after the patient underwent laser vaporization for Bowen’s disease, but finally responded to topical diflorasone. A biopsy of the center of the lesion prior to any attempted treatment was consistent with an erosion with edematous papillary dermis and a mixed cell perivasculitis without any description of a repaired epidermis nor any staining for specific cytokeratins.8

DISCUSSION: EMA OR PERSISTENT GRANULATION TISSUE?

When the non-healing surgical wounds described in the series by Jaffe et al6 and the case report by Mandrea8 are considered in aggregate, their clinical history is very similar in that all wounds demonstrate a failure to heal after prolonged periods ranging from three to six months, and their clinical description resembles that of persistent granulation tissue that fails to respond adequately to initial treatments, such as various combinations of silver nitrate, biologic dressings, topical corticosteroids, NSAIDs,6 or trichloroacetic acid.8 That any of these cases represent a unique complication of wound healing, namely that keratinocytes have failed to migrate versus inflammatory tissue has persisted and hindered migration, cannot be concluded from the clinical data presented. As far as the third article by Torne et al,7 neither the history nor the histological description are consistent with a persistent wound and this case report does not appear to be relevant at all.

The article by Jaffe et al6 gives the greatest detail of EMA as far as clinical history, histological work-up, and response to treatment. Nonetheless, the details of the authors’ methods for investigating and providing evidence for keratinocyte arrest are unclear and do not strongly support a unique clinical entity. First, as mentioned previously, the clinical descriptions are very consistent with chronic wounds that harbor persistent granulation tissue. Moreover, the location from where the histological specimens were obtained in the two cases was not specified, nor were the details of the exact histological findings provided. If the wounds were clinically consistent with granulation tissue and the biopsies obtained from the center of the wound, then it is unclear how this location would provide further data regarding the activity of the keratinocytes at the edge of the wounds, which must undergo epithelial to mesenchymal transition in order to undergo migration (see below). The authors do acknowledge that individual keratinocytes may have been lost in the processing of specimens, but keratinocytes appreciated on biopsy from the center of the wound would not provide representative data of the biochemical activity of keratinocytes at the edge of the wound. Secondly, the value of the cytokeratin stains performed on the wounds is unclear because this method was not detailed. The cytokeratin profile changes in a keratinocyte as it prepares to migrate because the keratinocytes undergo a transformation known as epithelial to mesenchymal transition (EMT).

EPITHELIAL TO MESENCHYMAL TRANSITION

The cytokeratin profiles of keratinocytes within a wound would reveal whether they are primed for migration or somehow arrested. Within 24 to 48 hours after epidermal injury, keratinocytes from wound edges migrate over a provisional irregular matrix comprised mostly of fibronectin and fibrin.9,10 Keratinocyte migration relies on critical interactions between the extracellular matrix (ECM) and, most commonly, integrin receptors on the cell surface of migrating keratinocytes.11 In order to induce and complete re-epithelialization, migrating keratinocytes undergo an EMT, in which they adapt a temporary and hybrid phenotype that allows them to migrate from the wound edges and appendages to infiltrate the wound bed.12 This modulation of keratinocyte cell-to-cell adhesion occurs in both physiological and pathological states, such as embryogenesis, cancer, and wound healing.

During EMT, keratinocytes downregulate keratins 1 and 10 in the suprabasal keratinocytes and upregulate keratins 6, 16, and 17.13 Within six hours after injury, there is a strong induction of keratins 6 and 16 in the keratinocytes at the wound edge that correlates with the reorganization of their keratin filaments from a pan-cytoplasmic distribution to a juxtanuclear location.14 The accumulation of keratin 16 has also been shown to coincide with the onset of re-epithelialization both spatially and temporally.15

Jaffe et al6 do not specify the cytokeratins for which they stained, and this detail would have confirmed whether the keratinocytes in these wounds had failed to prime themselves for migration and therefore arrested.

STEROIDS AND HYPERGRANULATION TISSUE

Although two studies6,8 cited the successful use of topical steroids with or without oral NSAIDs as having promoted epidermal maturation, potent topical steroids are also used to treat persistent granulation tissue. It has been shown that topical corticosteroids may benefit chronic wounds in terms of improving healing,16 pain relief,16 and treating persistent hypergranulation tissue.4,16,17 It is well-known that steroids have a sweeping effect on down-regulating cytokines at large, including initial inflammatory cytokines, such as interleukin-1 beta or tumor necrosis factor alpha that are found in acute wounds. However, the exception to these effects may be with regards to repair phase cytokines, such as transforming growth factor beta and platelet-derived growth factor, both of which may even be upregulated with steroid treatment.18 Essentially, topical corticosteroids have been reported to cause a clinically apparent decrease in hypergranulation tissue, thereby allowing epithelization to continue. This has been primarily reported anecdotally in the wound management literature.5

NSAIDS AND HYPERGRANULATION TISSUE

Similarly, NSAIDs have been shown to dampen persistent chronic inflammation. The arachadonic acid cascade involves a chain of enzymatic reactions that channel arachadonic acid, a fatty acid found in the phospholipids of cell membranes, into one of three enzymatic pathways: the cyclooxygenase (COX) pathway that generates prostaglandin derivatives, the lipoxygenase pathway that produces leukotrienes, and the cytochrome P450 dependant epoxidases. Human keratinocytes mostly utilize the cyclooxygenase and lipoxygenase pathways.19

As NSAIDs are COX inhibitors, the role of this pathway in wound healing has been investigated, but the data are insufficient to explain how COX inhibitors improve epithelialization. It has been shown that arachadonic acid release is necessary for integrin-mediated cell spreading, a marker for epithelialization, in murine fibroblasts.20 The metabolism of arachadonic acid by cyclooxygenase and lipooxygenase imparts different effects on fibroblasts. That is, lipooxygenase will transform arachadonic acid into protein kinase C epsilon that induces actin polymerization and cell spread.21 Arachadonic acid metabolized by COX produces prostaglandin E2, which leads to the bundling of actin filaments and possibly contributes to cell motility.22 As the importance of fibroblast motility via the action of COX in wound repair is essential, it would be logical that NSAIDs would inhibit wound repair; however, as reported in the case series, NSAIDs appeared to have improved wound healing.6 Moreover, studies have shown that neither nonselective COX nor selective (i.e., COX-2) inhibition have significant effects on the neovascularization, collagen deposition, or repaired tensile strength of acute incisional wounds in murine models.23 Within the first few days after wounding, wounds treated with either a selective or nonselective COX inhibitor did demonstrate subjectively fewer recruited neutrophils and macrophages, although this did not appear to render an appreciable effect on the morphology of the wounds or keratinocyte re-epithelialization, even at later time points.24

In one of the four cases described by Jaffe et al6, potent topical steroids incited re-epithelialization, whereas Case 2 required the addition of oral NSAIDs and Case 4 required the addition of an ultrapotent topical steroid after treatment with NSAIDs proved to be insufficient (Table 1). Interestingly, in Case 3, only NSAIDs were added when the wound was slow to heal, resulting in complete reepithelialization in two weeks time. In the case described by Mandrea,8 topical difluorasone proved to be adequate. As described above, topical corticosteroids are a commonly used treatment for persistent granulation tissue, but the case series by Jaffe et al6 is, to the authors’ knowledge, the first documenting the successful use of oral NSAIDS to treat a chronic wound whose clinical description is consistent with hypergranulation tissue. The authors extrapolated evidence for the efficacy of NSAIDS from the ophthalmology and dental literature that has reported, respectively, that NSAIDs do not affect corneal re-epithelialization and that prostaglandins can negatively affect periodontal wound healing.6

CONCLUSION

EMA has not been adequately supported with clinical and histological data. In our assessment, as supported by the above-mentioned studies and arguments, the cause of the observed delayed epithelialization is hypergranulation tissue. First, that the biopsies done by Jaffe et al6 were consistent with granulation tissue, a known surgical complication, is not surprising given their clinical presentation and response to topical corticosteroids. Similarly, the clinical description of the case by Mandrea et al8 and its response to topical steroids is also consistent with hypergranulation tissue. Secondly, the authors’ referral to their negative cytokeratin stain cannot be conclusive, as the exact concentration of the cytokeratins were not specified and therefore it is impossible to refute that these keratinocytes were unable to migrate. EMA is an interesting concept regarding the potential for an arrest in keratinocyte migration, although further studies are needed to better characterize the internal biochemical milieu of those keratinocytes at the edges of chronic non-healing surgical wounds.