Abstract: Bioengineered skin substitutes have emerged over the past 20 years as the most carefully studied and proven of the advanced wound management technologies. While the initial impetus for their development was to replace autograft, allograft, and xenograft in burn applications, they have found even wider application in the treatment of chronic venous and chronic diabetic ulcers. The current review addresses the history of skin substitutes, surveys the landscape of existing Food and Drug Administration-approved products and other promising innovations that appear close to market, and discusses the reasoning and controversies associated with design of these products. While acellular biologic constructs are discussed, the authors focus on products that include autologous or allogeneic cells. The various clinical trials supporting the use of skin substitutes for different wound healing indications are reviewed. The preponderance of literature supports the cost effectiveness of using skin substitutes in healing burn, autograft donor site, diabetic, and venous wounds. In addition, better methods for early identification of diabetic and venous ulcers that may not heal well with standard treatment should improve the process of triaging candidates for skin substitute therapy. In the future, attaining a more detailed understanding of the mechanisms by which skin substitutes induce accelerated healing, including a better appreciation of the roles of cytokines and cell scaffolds, may lead to product enhancements that increase efficacy. Ongoing progress toward overcoming issues, such as abbreviated shelf life and distribution difficulties, as well as high manufacturing costs, should enable broader implementation of skin substitutes in acute and chronic wound therapy.

Part 2 of 2

Continuing Issues in the Design of Cellular Skin Substitutes

There remain many controversies and unanswered questions associated with the design of cell-based wound therapies and cellular skin substitutes. For example, to what extent does addition of a neodermis improve results, especially in chronic wounds, compared with application of more simple cultured keratinocyte sheets? Fibroblasts are already plentiful in the beds of many chronic nonhealing wounds. Epidermal sheets seem to help regenerate dermis over time, and their engraftment requires attachment but no in-growth of blood vessels. Conversely do keratinocytes add anything to the action of a cultured fibroblast-containing matrix? Dermal replacement products ultimately become covered with a new epidermis. There is much evidence that dermal components may be effective in preventing wound contracture, which may be helpful in the cosmetic and functional outcome of healing in some circumstances. While the bilayered products (LSE, BCM) are elegant in design, in that, as co-cultures, they mimic skin more closely than a simple monolayer of keratinocytes, the relative importance of the artificial epidermis and dermis included in these devices is not yet fully established. Nevertheless, there are compelling in-vitro data, such as quantitative cytokine expression profiles (Figure 2), that imply a potentially broader capability for these products. Both LSE and BCM, for example, have shown significant efficacy in a wide range of acute and chronic wounds of varying etiology.[42]

Figure 2

This graph illustrates in-vitro cytokine expression by BCM (OrCel®) vs. LSE (Apligraf®): Cultures of equivalent-sized pieces of BCM (freshly thawed and rinsed in saline solution) and LSE (freshly removed from its packaging at room temperature) in BCM growth medium72 (Ortec International) were incubated for 48 hours, and media were sampled daily; assay of KGF-I and GMCSF were by commercial ELISA kits (Quantikine, R&D Systems) following manufacturer’s recommendations. Results are expressed as picograms (pg) of cytokine secreted per square centimeter area of tissue per day.

The matrix has been appreciated as important in providing a scaffold for cellular proliferation and migration, but its potential problems have been under-recognized. The poor durability of matrix collagen exposed to the enzyme-rich wound environment may limit the longevity and effectiveness of certain cell-based wound therapies and cellular skin substitutes. Once the matrix begins to disintegrate, the cells may wash away from the wound surface and their effect may be lost. BCM uses a cross-linked, partially protease-resistant collagen matrix to reduce this effect. The matrix also physically separates the epidermis from the wound bed; if keratinocyte-derived cytokines, for example, are of primary importance in the therapeutic effect of cellular skin substitutes, maintaining this diffusion distance may be less effective than applying the cells directly to the wound surface. Both LSE and BCM are relatively thin (<1mm) to attempt to minimize this effect.

Insofar as the actual mechanism of action of cell-based wound therapies is not completely known, the value of maintaining live cells that are capable of proliferating has actually not been demonstrated conclusively.[60] What has been shown is that metabolic activity, as distinct from proliferative capacity, is associated with wound healing efficacy.[21] Conversely, since the donor cells in a bilayered skin substitute applied to a venous ulcer appear to have very limited longevity,[15] there appears to be no safety concern regarding replicative competence of the cells.

Severe limitations on shelf life have led to various cryopreservation techniques that improve the logistics of product distribution and use. The extent to which the cryopreservation process and the thawing and rinsing procedures may damage cells within the product or wash away potentially beneficial growth factors remains an incompletely resolved question.[28] In the case of BCM, product recovered after thawing and rinsing has been shown capable of resuming apparently normal cell growth (not shown) and cytokine secretion (Figure 2) (Prajapati R, Silberklang M, manuscript in preparation), consistent with the absence of irreversible cell damage.

Clinical Results

Diabetic neuropathic wounds. Lipkin and colleagues61 have reported the results of a multicenter pilot trial of BCM for the treatment of neuropathic plantar wounds in 40 patients with diabetes mellitus. At the 12-week study end point, 7 of 20 treated wounds and 4 of 20 control wounds were completely reepithelized. Healing rate was nearly double (1.8% per day vs. 1.0% per day, p=0.008) in the treated compared with the control group, which received moist saline gauze and other standard wound care including debridement and offloading. The protocol, which prescribed up to six weekly applications of BCM, resulted in acceleration of wound closure that was highly statistically significant during the first four weeks before dropping closer to control levels. Benefits of the cellular skin substitute wound therapy were restricted to wounds smaller than 6cm2 (47% BCM patients healed vs. 23% control patients.) The FDA has granted approval for a pivotal trial of cryopreserved BCM in diabetic wounds.

Veves, et al.,[62] reported the results of a prospective, randomized multicenter trial of LSE in treating neuropathic diabetic plantar foot wounds that provided the basis for FDA approval of this product. The primary end point—complete wound closure at 12 weeks—was reached by 56 percent of LSE treated patients versus 38 percent of patients treated with standard wound care including debridement, offloading, and daily saline dressings. Median time to complete wound healing was lower in the LSE group (65 vs. 90 days, p=0.0026.)

The effectiveness of a dermal substitute (Dermagraft) in healing plantar foot wounds associated with diabetes mellitus was reported by Hanft and Surprenant.[38] Patients randomized to the dermal substitute treatment had up to eight applications of the bioengineered skin. At week 12, 71 percent of dermal substitute patients were healed compared with 14 percent of saline gauze-treated controls (p=0.004.) More recently, Marston and colleagues 39 reported a trial of the dermal substitute in 245 diabetic patients with plantar wounds. The incidence of complete healing was increased (30% vs. 18%; p=0.023) in the dermal substitute-treated wounds.

Venous leg ulcers. Very good results have been reported using bioengineered skin products to heal venous leg ulcers. Multiple weekly applications of LSE (average 3.3 per patient) plus compression induced complete healing in 63 percent of patients compared with 49 percent of those treated with compression alone (p=0.02).[63] Median time to healing was shorter (61 vs. 181 days, p=0.003) in the LSE group. The rate of wound recurrence at 12 months was not statistically different between study groups. The relative beneficial effect of this cellular skin substitute wound therapy was notable primarily in larger wounds and those of longer duration. Later analysis of patients meeting these criteria within the trial demonstrated that healing at six months was doubled using LSE.[64,65] The re-analysis of pivotal trial data on the effectiveness of LSE for hard-to-heal (greater than one year’s duration) venous wounds was reported by Falanga and Sabolinski in 1999.66 At six months, 47 percent of LSE-treated patients were completely healed compared with 19 percent of control patients (p< 0.005); median times to wound closure also showed a substantial advantage for LSE. Various subsequent reports from individual centers have confirmed similar good results using LSE on hard-to-heal ulcers of venous etiology.

BCM has also demonstrated excellent healing of venous wounds to date. In 2000, the pilot trial of BCM use in 36 venous ulcers showed that, compared with controls treated with foam dressings and standard compression, the product resulted in significantly more patients totally healed, by investigator assessment, at 12 (53% vs. 26%) and 26 (71% vs. 37%) weeks. Most patients received four applications of the skin substitute at weekly intervals. Wounds, as assessed by the investigators, were healed in a median 87 days with BCM, while more than half of control wounds were still unhealed at 180 days (p=0.033.) The efficacy of BCM was further supported by measurement of mean percent daily wound healing (2.0% for treated vs. 1.0% for control, p<0.0001.) Based upon these results, a pivotal trial using the cryopreserved form of BCM was initiated. Patient enrollment was completed in July, 2003. Preliminary analysis of the first 48 patients showed that 64 percent of BCM-treated patients had achieved complete wound closure by 12 weeks, compared with 39 percent of controls. The company is preparing its pre-market approval (PMA) submission as of January 2004.

Small series of venous ulcer patients treated with cultured epithelial allografts have been reported, noting good results,[48,67] but this technique has yet to be subjected to a prospective, multicenter, controlled, randomized trial.

Burns. Because of the life-threatening nature of deep, extensive burns and also because, in such situations, patients have limited sites for autologous skin graft donation, clinicians have had a low threshold for testing skin substitutes in this population. Many burn centers have tinkered with variations of CEA for wound coverage. Epicel has been used with excellent success, as noted above; typical “take” onto debrided burn wounds is 50 to 70 percent.[68]

In pilot, single-center studies, frozen cultured allogeneic keratinocyte sheets (Celaderm) have been noted to reduce healing time of partial-thickness burns by 44 percent (5.6 vs. 12.2 days), and were judged to obviate the need for autologous skin grafting in several of nine patients.[48] In a side-by-side comparison of cultured epithelial allograft applied to one section of skin graft donor sites and partial-thickness burns and control dressings applied to adjacent parts of the respective wound, healing was significantly improved in the cell-based wound therapy group (6.9 vs. 11.1 days, p<0.005.)[69] Severely burned patients treated with a regimen that included application of allograft and autograft epidermal cultures showed a survival advantage and reduced hospital stay compared with similar patients treated by other standard means at the same center.[70] Split-thickness donor sites for autologous graft harvest in burn patients have also been studied as suitable indications for intervention with skin substitute wound therapy. In a controlled matched-pairs design study of 82 burn patients, a single application of BCM was found to accelerate 100-percent epithelization of a donor site by an average of seven days, and with less scarring as compared to an acellular matrix product (Biobrane-L) control donor site in the same patient.[71] LSE has also been studied in this indication.[72]

Other indications. Wounds associated with several rare conditions have been successfully treated with cell-based wound therapies or skin substitutes. The original FDA-approved indication for BCM was in the surgical treatment of mitten hand deformity secondary to recessive dystrophic epidermolysis bullosa;[73] subsequent reports have indicated the utility of LSE in this condition as well.[74,75] Bullous morphea,[76] actinic purpura,[77] polyarteritis nodosa,[78] sarcoidosis,[79] corneal wounds,[80] and sickle cell disease[81] have reportedly been treated successfully with bioengineered skin. Another report cites both LSE and BCM as effective in accelerating healing of saphenous vein harvest wound complications.[82]

While the main focus in clinical trials of skin substitutes has been on the kinetics of wound closure (time to complete closure, rate of closure etc.), clinicians have also begun to consider more critically the quality of the healed wound. Waymack,[83] for example, used LSE over meshed split-thickness autologous skin grafts and noted that while controls healed as quickly as LSE-treated sites, those that had LSE demonstrated more normal pigmentation, vascularity, pliability and scar height after healing.

Economic/Reimbursement Issues

Published calculations of cost effectiveness of treatments are problematic and are often fraught with subjectivity. The cost of an unhealed wound may reach far beyond actual physician, hospital, and medical equipment bills and extend to foregone wages, diminished productivity by the employee, and additional direct medical costs for complications of nonhealing, such as infection. Further, some of the costs, while quite real, may be difficult or impossible to measure. Thus it is not surprising that various analyses of the cost-effectiveness of cell-based wound therapy or skin substitutes have reached conflicting conclusions.

Some authors are skeptical of high-cost treatments. Marston,84 for example, estimated the cost to heal small and very large venous ulcers to be about $1300 and $5300, respectively. Given this range, it is not obvious that a $1000 sheet of bio-engineered skin can show cost effectiveness. Harding and colleagues[85] performed a pooled analysis of 15 pressure ulcer and 12 leg ulcer studies and noted that relatively indiscriminate use of bioengineered skin products was quite inefficient. Sibbald modeled the healing and economic consequences of venous ulcer treatment with compression with or without one application of skin substitute;[86] his conclusion was that cost effectiveness was marginal, but was more pronounced in patients with ulcers of longer duration.

Other studies have reached exactly opposite conclusions and have stressed the economic benefits of cell-based wound therapy.[87] A review of 270 patients with venous ulcers included only the direct costs of care—physician visits, procedures, management of complications, dressings—and found that LSE-treated patients cost an average of $20,041 to heal compared with $27,493 for patients managed with Unna boot therapy.[88] Kirsner, et al.,[89] calculated that in their experience recalcitrant venous ulcers could be healed for approximately $16,000 using either LSE or by “standard” measures not including bioengineered skin products; the savings from faster healing and fewer complications were equaled by the cost of the cell-based wound therapy. Diabetic foot ulcers were treated less expensively using LSE compared with “standard” dressings in the Netherlands, and ulcer-free time was increased along with decreasing amputation incidence.[90] Nunez-Guttierrez and colleagues compared 39 severely burned patients treated by standard means with 32 similar patients whose regimen included the application of frozen cultured epithelial allograft.[70] Among the highest body surface area burns, patient survival was increased, and hospital stay decreased by over 20 percent as a result of quicker healing and fewer complications.

In the authors’ opinion, the bulk of evidence supports the contention that bioengineered skin products are cost-effective therapies for appropriately selected patients. The patients who would benefit the most are presumably those whose wound healing would be significantly delayed under standard therapy or whose wounds are most recalcitrant to standard therapy, and several researchers are attempting to define criteria by which to prospectively identify such patient subpopulations.[91–93]

Challenges and Future Directions

While currently available cell-based wound therapies are effective in accelerating wound repair and their appropriate use may achieve closure in nonhealing wounds, potential technologic improvements may further enhance the utility of this approach. For example, more effective cell preservation techniques could enhance shelf life and minimize issues related to storage. More specifically, simplified thawing and rinsing of cryopreserved products would make such products more user-friendly.

Studies need to more clearly define the optimal regimen of cell-based wound therapy (e.g., How many applications are useful? At what intervals should the therapy be applied?).

A more complete understanding of the mechanism of therapeutic action of bioengineered skin could lead to even more efficacious products. For example, if growth factor production is the essence of this modality, genetic modification of the cells to overproduce various specific cytokines might be feasible and productive. While preliminary efforts to create such products have been reported,[94] further refinements based upon advanced understanding may lead to substantial improvements.

Efforts by manufacturers to further reduce the cost of cellular skin substitute wound therapy could change the role of this approach dramatically. Rather than a later treatment to be used when other modalities have failed, or are likely to fail, bioengineered skin could be justifiably used earlier in the patient’s course. Lower cost could also allow for multiple applications and possibly increase the efficacy of the course of treatment. It is ironic that Medicare guidelines restrict the use of bioengineered skin to one application under most circumstances, while the clinical trials that have provided the basis for FDA approvals have generally used up to six or even eight weekly applications.

Eventually, the individual commercially available bioengineered skin products will be compared to each other, rather than to a non-cell based control treatment. Head-to-head comparison may establish winners and losers in this field, and these may be different for different indications.

Additional attention to the attributes of the healed wound could broaden appeal for the use of bioengineered skin if factors such as cosmesis, wound recurrence and scar thickness prove to be favorably influenced by its use.

Much additional work is required to better define the relative roles of the various wound technologies. In some instances, the beneficial effect of these may be additive; this has been reported, for example, for subatmospheric therapy (VAC®, Kinetic Concepts Inc., San Antonio, Texas) plus cell-based wound therapy.[95]

Published results implying excellent engraftment results using chimaeric syngeneic/allogeneic mixed cultured keratinocyte grafts deserve further investigation as potentially more rapid and cheaper clinical alternatives to pure cultured epithelial autografts.[96,97]

Building upon decades of advances in cell culture technology, cell-based, and cellular skin substitute wound therapy have recently come of age, and they represent the most empirically and scientifically proven of the modern wound care modalities. Certainly, incremental improvements upon current technologies will continue to appear. However, the authors believe that current use of bioengineered skin products is inappropriately low and that these products are too often reserved until patients have suffered pain and nonhealing for many months or even years. The advent of simple methods for early identification of patients who will not heal well with low technology treatment will improve the process of triaging candidates for bioengineered skin and will increase the popularity and cost-effectiveness of this approach.[90–92]