Proposed LCD - Skin Substitute Grafts/Cellular and Tissue-Based Products for the Treatment of Diabetic Foot Ulcers and Venous Leg Ulcers (DL39806) (2024)

Document Information

Coverage Guidance

Coverage Indications, Limitations, and/or Medical Necessity

Compliance with the provisions in this LCD may be monitored and addressed through post payment data analysis and subsequent medical review audits.

History/Background and General Information

Application of skin substitute graft and CTP for ulcer care indications other than DFU or VLU are not addressed by this LCD. Use of skin substitute graft/CTP must meet the reasonable and necessary threshold for coverage and these devices must be used in accordance with their intended use as approved/regulated by the United States (U.S.) Food and Drug Administration (FDA).

Depending on the purpose of the product and its proposed functions, skin substitute graft/CTP are regulated by the FDA premarket approval (PMA) process, FDA 510(k) premarket notification process, or the FDA regulations for human cells, tissues, and cellular and tissue-based products (HCT/Ps). A product with proposed benefit to chronic ulcer healing does not assume the designation of a skin substitute graft/CTP. FDA classification and indication are not the sole determinants of designation as a skin substitute graft/CTP or provide the reasonable and necessary threshold for coverage.

Chronic DFU and VLU may be unresponsive to initial therapy or persist despite appropriate standardized care. A DFU or VLU having failed to respond to standard of care (SOC) treatment after 4 weeks (28 days) may be considered chronic and the addition of a skin substitute graft/CTP may be considered reasonable and necessary for certain patients.^1-6

Patients receiving skin replacement surgery with a skin substitute graft/CTP should be under the care of a physician/non-physician practitioner (NPP) for the treatment of a systemic disease process (e.g., diabetes mellitus, chronic venous insufficiency, or peripheral vascular disease). It is imperative that systemic disease be monitored and treated to ensure adequate healing of the ulcer.^2,6,7

The medical record documentation must support the medical necessity for skin replacement therapy and the product’s use as an ulcer treatment, other than as a wound dressing or covering.

Covered Indications

If the patient meets all criteria as outlined in this LCD, application of a skin substitute graft or CTP in the treatment of DFU and VLU is considered reasonable and necessary for the following conditions:

1. The presence of a chronic, non-infected DFU having failed to respond to documented SOC treatment (outlined below) for a minimum of 4 weeks with documented compliance.^6-8
2. The presence of a chronic, non-infected VLU having failed to respond to documented SOC treatment (outlined below) for a minimum of 4 weeks with documented compliance.^4,6,9,10

For purposes of this LCD, SOC treatment includes^{4,5,7,9,11,12}:

• Comprehensive patient assessment (history, exam, Ankle-Brachial Index [ABI] and diagnostic tests as indicated) in an implemented treatment plan.
• For patients with a DFU: Assessment of Type 1 or Type 2 diabetes and management history with attention to certain comorbidities (e.g., vascular disease, neuropathy, osteomyelitis), review of current blood glucose levels/hemoglobin A1c (HbA1c), diet and nutritional status, activity level, physical exam that includes assessment of skin, ulcer, regional arterial perfusion (ABI), and assessment of off-loading device or use of appropriate footwear.
• For patients with a VLU: Assessment of clinical history (prior ulcers, phlebitis), physical exam (edema, skin changes), ABI, evaluation of superficial or deep venous reflux, perforator incompetence, and chronic (or acute) venous thrombosis. The use of a firm strength compression garment (>20 mmHg) or multi-layered compressive dressings is an essential component of SOC for venous stasis ulcers.^2,4,5

3. An implemented treatment plan demonstrating all the following^5,6,10:

• • Debridement as appropriate to a clean granular base
  • Documented evidence of offloading for DFU and some form of sustained compression dressings for VLU
  • Infection control with removal of foreign body or nidus of infection
  • Management of exudate with maintenance of a moist environment (moist saline gauze, other classic dressings, bioactive dressing, etc.)
  • Documentation of smoking history, and counseling on the effect of smoking on wound healing. Treatment for smoking cessation and outcome of counselling (if applicable)

4. The skin substitute graft/CTP is applied to an ulcer that has failed to heal or stalled in response to documented SOC treatment. Documentation of response requires measurements of the initial ulcer, pre- and post-completion of at least 4 weeks of SOC, with additional measurements at initial placement and each subsequent placement of the skin substitute graft/CTP. SOC measures without measurable signs of healing must have preceded the application for a minimum of 4 weeks and must continue for the course of therapy. Continuous compression therapy for VLU must be documented for the episode of care.^7,9,13

5. The medical record documentation must include the interventions that have failed during prior ulcer evaluation and management. The record must include an updated medication history, review of pertinent medical problems that may have arisen since the previous ulcer evaluation, and explanation of the planned skin replacement with choice of skin substitute graft/CTP product. The procedure risks and complications must also be reviewed and documented.^10-12,14,15

6. The patient is under the care of a qualified physician/NPP for the treatment of the systemic disease process(es) etiologic for the condition (e.g., venous insufficiency, diabetes, neuropathy) and documented in the medical record.^5,6,10,15

Coverage requirements for skin substitute grafts/CTP

To qualify as skin substitute graft/CTP the product must be:

1. A non-autologous human cellular or tissue product (e.g., dermal or epidermal, cellular and acellular, hom*ograft OR allograft), OR non-human cellular and tissue product (i.e., xenograft), OR biological product (synthetic or xenogeneic) which applied as a sheet, allowing the scaffolding for skin growth and is intended to remain on the recipient and grow in place or allow recipient’s cells to grow into the implanted graft material¹⁶ AND
2. Have quality supporting evidence to demonstrate the product’s safety, effectiveness, and positive clinical outcomes in the function as a graft for DFU and/or VLU.^4,10 Predicate products are not sufficient evidence for an individual product.

Note: Liquid or gel preparations are not considered grafts. Their fluidity does not allow graft placement and stabilization of the product on the wound.¹⁴

Limitations (per ulcer episode of care)

The following are considered not reasonable and necessary^2,4-6,9,17:

1. Greater than 4 applications of a skin substitute graft/CTP within the episode of skin replacement therapy (defined as 12 weeks from the first application of a skin substitute graft/CTP). In exceptional cases in which 4 applications is not sufficient for adequate wound healing, additional applications may be considered with documentation that includes progression of wound closure under current treatment plan and medical necessity for additional applications.¹⁷
2. Application of a skin substitute graft/CTP beyond 12-weeks per episode within the episode of skin replacement therapy. In exceptional cases in which 12 weeks is not sufficient for adequate wound healing, additional duration of care may be considered with documentation demonstrating progression of wound closure under current treatment plan and benefit expected from additional applications.
3. Repeat applications of skin substitute graft/CTP when a previous application was unsuccessful. Unsuccessful treatment is defined as increase in size or depth of an ulcer, no measurable change from baseline, and no sign of improvement or indication that improvement is likely (such as granulation, epithelialization, or progress towards closure). Unsuccessful therapy also includes reoccurrence of the ulcer in the same location within 12 months from initial application.
4. Application of skin substitute graft/CTP in patients with inadequate control of underlying conditions or exacerbating factors, or other contraindications (e.g., uncontrolled diabetes, active infection, progressive necrosis, active Charcot arthropathy of the ulcer extremity, active vasculitis, or ischemia)⁶
5. Use of surgical preparation services (e.g., debridement), in conjunction with routine, simple or repeat skin replacement surgery with a skin substitute graft/CTP
6. Excessive wastage (discarded amount)

• • The skin substitute graft/CTP must be used in an efficient manner utilizing the most appropriate size product available at the time of treatment. It is expected that use of product, size and preparation should conform to that most closely fitting the wound with the least amount of wastage.

7. All liquid or gel skin substitute products or CTPs for ulcer¹⁴
8. Placement of skin substitute graft/CTP on infected, ischemic, or necrotic wound bed^9,10

Provider Qualifications

Services provided within the LCD coverage indications will be considered reasonable and necessary when all aspects of care are within the scope of practice of the provider’s professional licensure. All procedures must be performed by appropriately trained providers in the appropriate setting.

Notice: Services performed for any given diagnosis must meet all the indications and limitations stated in this LCD, the general requirements for medical necessity as stated in CMS payment policy manuals, all existing CMS NCDs, and all Medicare payment rules.

Definitions

Autografts/tissue cultured autografts: Includes the harvest and application of an autologous skin graft. These products are designed to circumvent the challenges with autologous skin grafts in the treatment of chronic wounds, ulcers or burns.

Chronic Wound: A chronic wound may be defined as one that is physiologically impaired due to a disruption of the wound healing cycle as a result of impaired angiogenesis, innervation, or cellular migration, among other reasons for 4 weeks or longer.^3-5,18

Cellular and Tissue-Based Products (CTP) grafts (also called skin substitute graft*): Include hom*ologous human cellular and tissue products (e.g., dermal or epidermal, cellular and acellular, hom*ograft or allograft), non-human cellular and tissue products (i.e., xenograft), and biological products (synthetic or xenogeneic) that form a sheet scaffolding for skin growth when applied in a sheet over an open wound or ulcer to augment closure or skin growth.^2,16

Failed response: An ulcer that has increased in size or depth, or no change in baseline size or depth, or no sign of improvement or indication that improvement is likely (such as granulation, epithelialization, or progress towards closing).

Stalled Wound: An ulcer that has entered a nonhealing or intransigent phase.¹⁹

Wound dressing or coverings: Applications applied to wounds as a selective barrier to clean, cover and protect wounds from the surrounding environment to promote optimal environment for wound healing.

*There is a lack of clarity in the definition of skin substitute. For the purpose of this LCD, skin substitute grafts will align with the AMA CPT^® codebook¹⁶ description “non-human skin substitute grafts and biological products that form a sheet scaffolding for skin growth”. This surface is not intended to be removed, but grows into place or serves as surface for new skin to grow in.

Summary of Evidence

A literature search was conducted using the following key words: non-healing; ulcer; chronic; diabetic foot; foot ulcer; venous leg ulcer; guidelines; ulcer healing; skin substitutes; dermal skin substitute; human skin allograft (HSA); randomized trial; SOC; venous leg ulcer; skin grafts; ulcer dressing; human derived products; animal derived products; FDA regulations. The literature search was filtered to locate articles within 5-10 years, full-text articles, clinical trials, and systematic reviews. In general, improved health outcomes of interest include patient quality of life and function.

Case reports, case series and retrospective reports were not reviewed due to low quality evidence. Review papers, editorials and unpublished reports were not included in the analysis. Literature for use of products outside of DFU and VLU was not included in the evidence review. Much of the literature for DFU overlapped with diabetic lower extremity ulcers so both are included in the review.

Introduction

DFU and VLU, having failed to respond to 4 weeks SOC treatment, may meet reasonable and necessary criteria for the adjunctive application of a skin substitute graft/CTP for certain patients. DFU may affect up to 6% of Medicare beneficiaries with either Type I or Type II diabetes. Chronic wounds such as DFU and VLU impact patient quality of life due to impaired mobility, pain, and progressive morbidity.^2,4,6,20

Evidence-Based Guidelines for SOC

Evidence-based guidelines indicate that SOC treatment of lower extremity ulcers (e.g., DFU and VLU) should include mechanical offloading, infection control, mechanical compression, limb elevation, debridement of necrotic tissue, management of systemic disease and medications, nutrition assessment, tissue perfusion and oxygenation, education regarding care of the foot, callus, and nail and fitting of shoes, as well as counseling on the risk of continued tobacco use. In addition, maintenance of a moist ulcer environment through appropriate dressings facilitates development of healthy granulation tissue and epithelialization and potentiates complete healing at an ulcer site. Dressings are an integral part of ulcer management by maintaining a moist environment, limiting contamination, and absorbing exudate.^{5-7,9,11,12,15}

A comprehensive assessment of patients and their ulcers will also facilitate appropriate care by identifying and correcting systemic causes of impaired healing. The presence of a severe illness or systemic disease and drug treatments, such as immunosuppressive drugs and systemic steroids, may inhibit ulcer healing by changes in immune tolerance, metabolism, inflammation, nutrition, and tissue perfusion. Therefore, this information, in conjunction with a detailed history of the ulcer itself, is essential.^6,7,9

Vascular evaluation is also vital for all chronic ulcers. Palpation of pulses may be problematic in cases of medial arterial calcification. An ABI should be taken for patients with a questionable pulse deficit although the ABI levels may be falsely elevated with medial arterial calcification. The patient is considered to have impaired arterial perfusion when the ABI is below 0.9. Any ABI less than 1.0 suggests a degree of vascular disease.⁵ To supplement ankle-brachial studies, toe blood pressure readings, pulse volume recordings, transcutaneous oxygen measurements (TCOMs), and skin perfusion pressure measurements have been suggested as acceptable benchmarks for the prediction of ulcer healing.

Venous ulcers require a series of diagnostic testing to verify superficial or deep venous reflux, perforator incompetence, and chronic (or acute) venous thrombosis. In this regard, venous duplex ultrasound is recommended and if the venous duplex ultrasound does not provide definitive diagnostic information, a venous plethysmography is recommended. Patients with mixed arterial and venous disease require a combination of arterial and venous noninvasive testing. The use of a Class 3 (most supportive) high-compression method is strongly recommended in the treatment of venous ulcers. High strength compression may be applied using techniques such as multilayered elastic compression, inelastic compression, Unna boots, compression stockings, and others. The extent of compression should be modified for patients with mixed venous/arterial disease.^5,8,9

The clinical practice guidelines of the Society for Vascular Surgery and the American Venous Forum recommend that patients with VLU have the ulcer classified using the Clinical class, Etiology, Anatomy, and Pathophysiology (CEAP) classification (confirmed by duplex scan). The Venous Clinical Severity Score (VCSS) is recommended to assess changes in response to therapy. Specific classification of venous disease is essential for standardization of venous disease severity and evaluation of treatment efficiency.⁹

The Society for Vascular Surgery in collaboration with the American Podiatric Medical Association and the Society for Vascular Medicine has recommended a SOC treatment schedule for DFU includes weekly to monthly ulcer evaluations of size and healing progress, infection control, debridement of all devitalized tissue and surrounding callus material, dressings that maintain a moist ulcer environment, control of exudate, and avoidance of maceration of adjacent intact skin. Adequate glycemic control is also recommended to reduce the incidence of DFU and infections in addition to periodic assessments of appropriate footwear or off-loading devices.^7,9,13

Evidence-Based Guidelines for Skin Substitute Grafts/CTP

Skin substitute grafts/CTP are a heterogeneous group of biological and synthetic elements that allow the temporary or permanent occlusion of ulcers. Dermal substitutes may vary from skin xenografts or allografts to a combination of autologous keratinocytes over the dermal matrix, but all address the goal of resemblance to an individual’s skin to the greatest extent possible.¹³ Skin substitute grafts/CTP are considered an advanced therapy adjunctive to the established SOC treatment protocols for ulcer care to increase the chances of healing. In this regard, evidence-based guidelines recommend ulcer bed preparation prior to the application of any biologically active dressing which includes complete removal of slough, debris, and necrotic tissue.¹⁵ Skin substitute grafts may be considered in conjunction with SOC treatment for DFU that have failed to demonstrate more than 50% ulcer area reduction after a minimum of 4 weeks of standard therapy.⁷ For VLU, if substantial ulcer improvement is not demonstrated after a minimum of 4 weeks of standard therapy, skin substitute grafts or CTP may be considered in addition to compression therapy.^4,6,9

Product Classifications

Several classification systems have been proposed to categorize products, but there is not a universally accepted classification system. The products vary widely, ranging from synthetic or natural, a variety of origins, with additives and processing impacting the final product.²¹ Even products derived from the same origin are variable since these products undergo proprietary processing. Skin substitute grafts/CTP may share similarities, but they are individually unique in their proprietary processing, thickness, cell count, presence of living cells and other features. In 2001, a classification system was proposed that put skin substitutes into 3 groups: Class I cultured epidermal equivalents alone, Class II dermal components which are from processed skin or have been manufactured with extracellular matrix (ECM) proteins, and Class III both dermal and epidermal.²² Kumar created a classification system in 2008 which divides products into 3 classes based on temporary, single and bilayer materials and included dermal or epidural and natural or synthetic into the classification.²³ The Davison-Kotler classification system was developed to classify the differences between the products based on functionality according to cellularity (acellular or cellular), layering (single or bilayer), replaced region (epidermis, dermis, or both), material used (natural, synthetic, or both), and permanence (temporary or permanent).²² The result is products within the same class, varying significantly and impacting the product’s function indeterminately.²⁴

Potential Harm

The potential harm of skin substitutes is challenged by lack of high-quality studies and long-term data. The risk of human-based products include infections being transmitted from the donor tissue to the recipient. Most products undergo stringent processing to reduce this risk, but bacterial and viral transmission risk remains. The duration of cells delivered and effect in the wound basem*nt membrane is not fully understood.²⁵ Some types of grafts are at risk of graft rejection and there is variability in cosmetic results. Adherence to underlying tissues may vary based on hydrophilic surface properties of the graft which may impact effectiveness.²⁵ Concerns have arisen regarding specific constituent molecules within the matrix with the potential to elicit adverse responses in host tissues. The mechanism of changes in the ECM through cell-matrix interactions and ECM remodeling is not fully understood, eliciting concern for the derived microenvironment promoting tumorigenesis, metastasis, inflammatory or autoimmune disease evolution.²⁶ Very few studies explore long term safety of skin substitute grafts/CTP so true risk associated with these products remains unclear.

Health Care Disparities

There is a paucity of literature addressing health care disparities in the use of skin substitutes specifically for DFU and VLU. Diabetic management is known to be impacted by social determinants of health with worse outcomes noted in minority and socioeconomically disadvantaged populations. Comprehensive care models with multidisciplinary teams have proven effective in treatment of DFU by improving access to care, access to specialist and effective and timely treatment.²⁷ Teams include a combination of primary care, endocrinology, vascular surgeons, orthopedic surgeons, podiatrists, and wound care specialists. The literature reviewed for DFU included patients with diabetes. The majority of reviewed literature did not represent racial diversity with subjects outside the Medicare population. Future research should aim to include a diverse population representative of those impacted by the condition and include representation of the Medicare population in age distribution.

Agency for Healthcare Research and Quality (AHRQ) Technical Brief

The AHRQ provided an evidenced-based technical brief for skin substitute grafts for treating chronic ulcers.⁴ This technical brief was developed to describe assorted products that may be considered skin substitute grafts in the U.S., which are utilized for the treatment of chronic ulcers. In addition, systems utilized to classify skin substitute grafts were assessed, randomized controlled trials (RCTs) involving skin substitute grafts were reviewed, and recommendations were made regarding best practices for future studies. Search of the published literature since 2012 was conducted for systematic reviews/meta-analyses, RCTs, and prospective non-randomized comparative analysis studying commercially available skin substitute grafts for individuals with DFU, VLU, pressure ulcers, and arterial leg ulcers.

Seventy-six skin substitute grafts were identified and categorized using the Davison-Kotler classification system, a method structured according to cellularity, layering, replaced region, material used, and permanence. Of these, 68 (89%) were categorized as acellular dermal substitutes, largely replacements from human placental membranes and animal tissue sources. Acellular dermal substitutes prepared from natural biological materials are the most common commercially available skin substitute graft products for treating or managing chronic ulcers. Cellularity is a significant difference among skin substitute grafts as the presence of cells raises the rejection risk and production complexity. This category includes decellularized donated human dermis (14 products recognized), human placental membranes (28 products recognized), and animal tissue (21 products recognized). Fewer products are prepared from synthetic materials (2 products recognized) or a blend of natural and synthetic materials (2 products recognized). A limited number of skin substitute products are acellular replacements for both the epidermis and dermis (1 product recognized). Only 8 products were recognized that contained cells and would be classified in the cellular grouping.

Three systematic reviews and 22 RCTs studied the utilization of 16 distinct skin substitutes, comprising acellular dermal substitutes, cellular dermal substitutes, and cellular epidermal and dermal substitutes in DFU, pressure ulcers, and VLU. Twenty-one ongoing studies (all RCTs) assessed an additional 9 skin substitute grafts with comparable classifications. It was noted that studies seldom reported clinical outcomes, such as amputation, ulcer recurrence at least 2 weeks after treatment ended, or patient-related outcomes, such as return to function, pain, exudate, and odor. This review found that more studies are needed to assess the effectiveness of most skin substitutes and this future research needs to be better designed and include clinically relevant outcomes.

Of the 22 included RCTs, 16 studies contrasted a skin substitute with SOC. The SOC for each ulcer type involved sharp debridement, glucose control, compression bandages for VLU, pressure redistribution support surfaces for pressure ulcers, infection control, offloading, and daily dressing changes with a moisture-retentive dressing, such as an alginate or hydrocolloid type dressing. Though 85% of the studies examining acellular dermal substitutes portrayed the experimental intervention as favorable over SOC for ulcer healing and quicker time to heal, the data is not adequate to determine whether ulcer recurrence or other sequela are less frequent with acellular dermal substitutes. Only 3 studies contrasted cellular dermal substitutes with SOC. Clinical evidence for cellular dermal substitutes may be limited by the lack of robust, well-controlled clinical trials.

Of the 6 head-to-head comparative studies, results from 5 studies did not show substantial differences between skin substitute grafts in outcomes measured at the latest follow-up (>12 weeks). One study concluding at 12 weeks described a substantial difference in ulcer healing favoring an acellular dermal skin substitute over a cellular epidermal and dermal skin substitute. Another study compared 2 acellular dermal substitutes and seemed to have deliberately underpowered 1 arm of the study as the statistical significance was not elucidated or expected for this study arm. Of the 2 studies reporting on recurrence, 1 study described comparable recurrence, while the other study reported no recurrence at 26 weeks. The current evidence base, as portrayed by the authors for the literature reviewed, may be inadequate to determine superiority of 1 skin substitute graft product over another.

The report acknowledges the potential risk of bias due to 20 of the 22 RCTs reviewed being industry sponsored. This AHRQ Technical Brief also noted that a skin substitute’s commercial availability is not a reflection of its legal status. Manufacturers self-determine whether their HCT/P may be marketed without FDA preapproval and frequently misunderstand or mischaracterize the conditions necessary for the product to be regulated solely for communicable disease risk. The Code of Federal Regulations 21 CFR 1271.10(a) is referenced; FDA Announces Comprehensive Regenerative Medicine Policy Framework was cited.¹⁴

Systemic Review and Meta-Analysis

Santema et al.²⁸ provided a systematic review and meta-analysis to assess the efficiency of skin substitute grafts utilized for the treatment of DFU regarding ulcer healing and limb salvage. Using the Cochrane Collaboration methodology, 17 clinical trials were identified, which included a total of 1,655 randomized study participants with DFU. The number of study participants per clinical trial ranged from 23 to 314. Fourteen studies included chronic or difficult to heal ulcers that were present for a minimum of 2, 4, or 6 weeks.

Skin substitute grafts were contrasted with SOC in 13 trials. The results collectively demonstrated that SOC treatment, combined with a skin substitute product enhanced the chances of attaining complete ulcer closure in contrast to SOC alone after 6 to 16 weeks (risk ratio [RR] 1.55, 95% confidence interval [CI] 1.30 to 1.85, low quality of evidence). Apligraf^®/Graftskin^®, EpiFix^®, and Hyalograft 3D™ were the only individual products that demonstrated a statistically substantial beneficial effect on complete ulcer closure (i.e., full epithelialization without any evidence of drainage or bleeding). Four clinical trials contrasted 2 different types of skin substitutes, although no product demonstrated a greater effect over another. Sixteen of the trials evaluated the efficacy of a bioengineered skin substitute. Only 1 trial evaluated the efficacy of a non-bioengineered skin graft.

The total occurrence of lower limb amputations was only reported for 2 trials and the results for these 2 trials collectively produced a substantially lower amputation rate for individuals treated with skin substitute grafts (RR 0.42 95% CI 0.23 to 0.81), though the absolute risk difference (RD) was small (-0.06, 95% CI -0.10 to -0.01, very low quality of evidence). Of the included studies, 16 reported on adverse events (AEs) in diverse ways, although there were no reports of a substantial difference in the incidence of AEs between the intervention and the control group. Additionally, support of long-term effectiveness is lacking, and cost-effectiveness is unclear. Noted limitations included a variable risk of bias among the studies, the lack of blinding (i.e., study participants and investigators knew which patients were receiving the experimental therapy and which patients were receiving the standard therapy), and 15 of the studies conveyed industry involvement; the majority did not indicate if the industry applied any limitations regarding data analysis or publication.²⁸

Jones et al.²⁹ systematic literature review sought to evaluate the effect of skin substitute grafts for the treatment of VLU. Using the Cochrane Collaboration methodology, 1 new trial was identified, generating a total of 17 RCTs, which included a total of 1,034 study participants. The studies were comprised of participants of any age, in any care setting with VLU. Given that the process for diagnosis of venous ulceration differed between studies, a standard definition was not applied. The trials also involved study participants with arterial, mixed, neuropathic, and diabetic ulcers provided that the outcomes for patients with venous ulcers were conveyed separately. To be included in the review, trials also had to report at least 1 of the primary outcomes objective measures of healing, e.g., relative or absolute rate of change in ulcer area, time for complete healing, or proportion of ulcers healed within the trial period.

Eleven studies contrasted a graft with SOC. Two of these studies (102 patients) contrasted an autograft with a dressing, 3 studies (80 patients) contrasted a frozen allograft with a dressing, and 2 studies (45 patients) contrasted a fresh allograft with a dressing. Two studies (345 patients) contrasted a tissue-engineered skin (bilayer artificial skin) with a dressing. In 2 studies (97 patients), a single-layer dermal replacement was compared with SOC.

Six studies compare alternative skin grafting techniques. The first study (92 patients) differentiated an autograft with a frozen allograft; a second study (51 patients) contrasted a pinch graft (autograft) with a porcine dermis (xenograft); the third study (110 patients) compared growth-arrested human keratinocytes and fibroblasts with a placebo; the fourth study (10 patients) analyzed an autograft delivered on porcine pads with an autograft delivered on porcine gelatin microbeads; the fifth study (92 patients) contrasted a meshed graft with a cultured keratinocyte autograft; and the sixth study (50 patients) contrasted a frozen keratinocyte allograft with a lyophilized (freeze-dried) keratinocyte allograft.

Overall, the results show that substantially more ulcers healed when treated with bilayer artificial skin than with dressings. There was inadequate evidence from the other trials to establish whether other types of skin grafting improved the healing of venous ulcers. The authors concluded that bilayer artificial skin, used together with compression bandaging, improves venous ulcer healing as compared to a simple dressing plus compression.

It was noted that the overall quality of the studies reviewed was poor, thus affecting the risk of inherent bias. Many studies did not have inclusion criteria or insufficient information regarding randomization techniques. In addition, withdrawals and AEs were inadequately reported. Deficient data regarding withdrawals and the inclination to perform per-protocol analyses rather than intention-to-treat (ITT) analyses signify that the outcomes in the original study documentation may be biased.²⁹

A 2017 meta-analysis of RCT comparing amniotic tissue products to SOC in nonhealing DFU was conducted. PubMed^®, Cochrane Central Register of Controlled Trials, and the Cochrane Database of Systematic Reviews search identified 596 potentially relevant articles of which 5 met the selection criteria. The pooled set included 259 patients and the pooled relative risk of healing with amniotic products compared with control was 2.7496 (2.05725–3.66524, p< 0.001). The products included in this analysis were AmnioExcel^®, EpiFix^®, and Grafix^®. Four trials changed the amniotic product weekly; 1 paper reported an average of 2.5 applications of EpiFix^®, and in 1 study where reapplication was at the discretion of the clinician, no decrease in healing was found compared to the per protocol application changes. The author concludes that there is benefit in healing rates of amniotic products for DFU and if this impacts other outcomes and subsequent complications such as amputation and death, further investigation will be required.³⁰

A 2020 systematic review/meta-analysis reported on complete healing rates for DFU with acellular matrix.³¹ Nine RCT with 897 patients were included. They report those treated with an acellular matrix had higher healing rates at 12 weeks (RR = 1:73, 95% CI: 1.31 to 2.30) and 16 weeks (RR = 1:56, 95% CI: 1.28 to 1.91), a shorter time to complete healing (mean difference [MD] = −2:41; 95% CI: -3.49 to -1.32), and fewer AEs (RR = 0:64, 95% CI: 0.44 to 0.93) compared to SOC. RCTs include Graftjacket™, OASIS^® ULTRA Matrix, DermACELL^®, Integra^® and AlloPatch^®. The heterogenicity reported varies depending on the outcome measures but the analysis is limited by high variety in wound types, differing products and number of applications, variations in SOC in control arms, different durations of treatment and risk of bias in the included studies.

A 2017 systematic review and meta-analysis identified 6 RCTs comparing acellular dermal matrix (ADM) to SOC for DFU. Different commercial products of ADM were included in this meta-analysis, including DermACELL^®, Graftjacket™, Integra^® Dermal Regeneration Template (IDRT), and human reticular acellular dermis matrix (HR-ADM). The pooled group included a total of 632 DFU patients and sample size of the studies ranged from 14-154 for a duration of 4-16 weeks. Studies were pooled and analyzed (with and without the study that only extended to 4 weeks) and concluded that complete healing rate in the ADM group was higher than SOC [RR 2.31, 95% CI 1.42 to 3.76 I²=74%] which is 2.31 times more likely for complete wound healing than SOC at 12 weeks. The authors rated the strength of evidence as moderate and acknowledge the limitation related to lack of blinding. This meta-analysis is limited by risk of publication bias, lack of uniform ADM the products, variation in dressing products and potential variation within SOC, different application number amongst the different studies, small samples of individual studies, short term follow up, and fairly high heterogeneity within some outcome measures leading to the authors call for more robust studies.³²

A 2012 systematic review of RCTs evaluating wound closure rates for patients treated with advanced wound matrix compared to SOC for VLU was conducted.³³ One RCT was found for 3 products Apligraf^® [n = 130 treatment, n = 110 control]; OASIS^® Wound Matrix [n = 62 treatment, n = 58 control]; and Talymed^® [n = 22 treatment, n = 20 control]. In the Talymed^® study 62 patients in the treatment arm with varying applications frequency reported statistically significant closure rate compared to SOC, but this was only found in 1 of the 3 arms (biweekly application group). Risk of bias assessment was not conducted, but they reference the AHRQ report⁴ which reported higher degree of bias for the Apligraf^® and OASIS^® studies that were included due to lack of blinding. Limitations of this report include variations in assessment period across the studies, baseline wound characteristics were not compared, and only 1 arm of the Talymed^® study was included with a sample size too small to determine effect.

Systematic reviews for skin substitute graft/CTP are challenged by multiple factors. These reviews pool different products with different features, types of wounds, baseline health factors, duration of treatment, number of applications and variations in SOC creating significant factor variance. Even within the same study variability in SOC and managements are not clearly defined. Most included studies have notable risk of bias, small sample sizes, and short-term follow-up resulting in overall low-quality literature. The systematic reviews are limited by the quality of the included studies and heterogeneity between studies so even with positive outcomes, there is a lack certainty that the effect is due to the skin substitute/CPT itself.

Clinical Trials for Skin Substitute Grafts or CTP for DFU

Affinity^®

A multicentered, RCT was conducted across 14 centers to assess the clinical outcomes of hypothermically stored amniotic membrane (HSAM) vs SOC for DFU. After a 2-week screening phase, 76 participants were randomized with random allocation sequence to either Affinity^® or SOC and followed for 16 weeks. Wound measurements were validated with an Aranz laser-assisted wound measurement device. Product was changed weekly or until the ulcer healed. Wound closure for Affinity^®-treated ulcers (n=38) was significantly greater than SOC (n=38) by 12 weeks (55 vs 29%; p = 0.02) and 16 weeks (58 vs 29%; p = 0.01) respectively.³⁴ Strengths of the study include the randomization, screening and follow-up phase, comparison to SOC, adequately powered, the use of multi-centered sites and an overall low risk of bias. Limitations include lack of blinding, and short-term follow-up. This study also addresses a population with complex wounds extending into the muscle or tendon which is a particularly difficult population to treat. Additional studies include a basic science report exploring the mechanism of how the product may work,³⁵ and a case report and series.

AlloPatch^®/Flex HD^®/Allopath HD^®/Matrix HD^®

Zelen et al.³⁶ performed a RCT to evaluate the healing rates, safety, and cost using an open-structure human reticular acellular dermal matrix (HR-ADM) (i.e., AlloPatch^® Pliable⁾ plus SOC to SOC alone for DFU. A total of 40 subjects were randomized to HR-ADM plus SOC (n = 20) (AlloPatch^® applications weekly) or SOC alone (n = 20). The primary outcome of this study focused on a comparison of ulcer healing at 6 weeks between these 2 groups. Wounds were considered as healed if there was complete (100%) re-epithelization with no drainage and no need for dressing. At 12 weeks, 80% (16/20) of the AlloPatch^®-treated ulcers had healed contrasted with 20% (4/20) of the ulcers treated with SOC alone (p=0.00036). The mean time to heal within 12 weeks was 40 days (95% CI: 27–52 days) for the AlloPatch^® vs 77 days (95% CI: 70–84 days) for the SOC group (p=0.00014). The average number of AlloPatch^® grafts used to achieve closure per ulcer was 4.7 (SD=3.3) at 12 weeks. There was no occurrence of increased AEs or serious adverse events (SAEs) between groups, or any AEs related to the graft. This study concluded that the use of AlloPatch^® plus SOC is more effective in the treatment of DFU than with SOC alone. However, this study had high risk of bias due to missing outcome data and was also limited by short-term follow-up, and small sample size. The authors also followed the patients that failed SOC arm and were eligible for cross-over treatment in a retrospective format. Twelve patients received the allograft and 83% achieved complete wound healing with mean time of 21 days to closure.³⁷ Due to the retrospective study design it is not clear if the wounds would have closed with continued SOC during this timeframe.

Literature was also found in breast reconstruction, rotator cuff repair, hernia repair, and lab research.^38-40

AmnioBand^®

Glat et al.⁴¹ conducted a RCT to contrast a dehydrated human amnion and chorion allograft (dHACA) (i.e., AmnioBand^®) with SOC and a tissue-engineered skin substitute (TESS) (i.e., Apligraf^®) with SOC in the treatment of DFU. At the 12-week assessment, it was found the mean time to healing was 32 days. (95% CI, 22.3–41.0) for the AmnioBand^® group vs 63 days (95% CI, 54.1–72.6) for the Apligraf^® group. The healing rate at 12 weeks was 90% (27/30) for the AmnioBand^® group vs 40% (12/30) for the Apligraf^® group. Limitations noted for this study include lack of blinding, short-term follow-up, and high risk of bias.⁴¹

DiDomenico et al.⁴² conducted a prospective, RCT to compare a dHACA (i.e., AmnioBand^®) used with SOC to SOC alone in the treatment of DFU for up to 12 weeks. At 6 weeks, 70% (14/20) of the DFU in the AmnioBand^® group achieved healing compared to 15% (3/20) of the DFU in the SOC group. At 12 weeks, 85% (17/20) of the DFU in the AmnioBand^® group healed compared with 25% (5/20) in the SOC group (mean time to heal of 36 and 70 days, respectively). At 12 weeks, the average number of grafts used per healed wound for the AmnioBand^® group was 3.8 ± SD 2.2 (median 3.0). All analyses used the ITT approach, and the risk of bias was low. Limitations were short-term follow-up and lack of blinding.⁴²

DiDomenico et al.⁴³ performed a RCT to compare a dHACA (i.e., AmnioBand^®) used with SOC to SOC alone in the treatment of DFU for up to 12 weeks. Eighty patients participated in the study: 40 patients in the AmnioBand^® group and 40 patients in the SOC group. The AmnioBand^® was applied weekly during the study period until healing occurred (complete epithelialization without drainage), the patient was withdrawn, or the study was completed. At 6 weeks, 68% (27/40) of the DFU in the AmnioBand^® group achieved healing contrasted to 20% (8/40) of the DFU in the SOC group (p=1.9 × 10⁻⁵). At 12 weeks, 85% (34/40) of the DFU in the AmnioBand^® group achieved healing compared with 33% (13/40) of the DFU in the SOC group. The average time to heal within 12 weeks was substantially quicker for the AmnioBand^® group contrasted with the SOC group, 37 days vs 67 days in the SOC group (p=0.000006). The average number of grafts used per healed wound during the same time was 4.0 (SD: 2.56) at 12 weeks. All analyses used the ITT approach, and the risk of bias was low. Limitations include lack of blinding, and short-term follow up.

Amnioband^® is also reviewed in the VLU section.

AmnioExcel^®

Snyder et al.⁴⁴ performed a multi-center RCT for assessment of a dehydrated amniotic membrane allograft (DAMA) (i.e., AmnioExcel^®) with SOC (n=15) in comparison to SOC alone (n=14) for chronic DFU for 6 weeks. The AmnioExcel^® with SOC group wounds were debrided, AmnioExcel^® applied, covered with non-adherent dressings, lightly secured, and wrapped with a compression dressing. Patients in the AmnioExcel^® with SOC group had a total of 4.3 + 1.7 allografts applied; frequency of the application was left to individual provider. Results showed that 33% of patients in the AmnioExcel^® with SOC group achieved complete wound closure at or before week 6, compared with 0% of the SOC alone group (ITT population, p=0.017). The per protocol population showed 45.5% of patients in the AmnioExcel^® with SOC group achieved complete wound closure, while 0% of SOC-alone patients achieved complete closure (p=0.0083). Limitations of this study included 4 early withdrawals leaving only 25 patients in the final cohort, small sample size, lack of blinding, and high risk of bias due to stratification of wound type prior to randomization, per-protocol reporting only without intention to treat analysis, and lack of validation of outcome measurements. The authors call for the need for additional studies which are necessary to confirm if the findings were related to the allograft and longer-term follow-up.⁴⁴

Apligraf^® (formerly Graftskin^®)

A prospective RCT was comprised of patients with DFU. A total of 208 patients from 24 sites were randomly assigned into either the Apligraf^® (formerly Graftskin^®) (n=112 patients) or SOC (n=96 patients) group and followed for 12 weeks. Complete wound healing was reported in 56% of Graftskin^® patients compared to 38% of the control group. Authors report Graftskin^® time to complete closure was significantly lower than the SOC group (p=0.0026). Forty-four patients withdrew from the study before study completion. Average applications of Graftskin^® per patient were 3.9 (range 1-5) for the duration of the study. The average number of applications of 3.9 (range 1-5) with 1 application [n=10]), 2 applications [n=11], 3 applications [n=15], 4 applications [n=17] and 5 applications [n=59]) used per patient over 12 weeks. Ulcer recurrence was 5.9% in the Graftskin^® group and 12.9% in the control group at 6 month follow up. Limitations include high risk of bias, moderate number of patients lost to follow up, and additional dressing changes allowed in both groups if ulcer was not healed by week 5.⁴⁵

A prospective, multicenter, open-label RCT compared Apligraf^® plus SOC to SOC alone in DFU patients in the European Union (EU) and Australia to a similar study in the U.S. The EU and Australian studies were comparable and data from both studies were pooled. The EU and Australian studies were comprised of 72 patients, 33 in the Apligraf^® group and 29 in SOC group and the U.S. study was compromised of 208 patients: 112 in Apligraf^® group and 96 in the control group. The mean ulcer duration was significantly longer in the EU and Australian study (21 months) compared to 10 months in the U.S. AEs were reported for 12 weeks in both studies and were comparable and not related to the graft. At 12 weeks, combining the data from both studies, 55.2% of the Apligraf^® group achieved wound closure as compared to 34.3% in SOC arm (P = 0.0005; Fishers exact test), and Apligraf^® subjects had a significantly shorter time to complete wound closure (P = 0.0004; log-rank test). Limitations include premature study closure (non-safety related) for the EU and Australian studies which were underpowered due to halting study enrollment. Due to pooling of 2 different studies, it was difficult to assess risk of bias of the individual studies.⁴⁶

An international multi-center, RCT was conducted in patients with DFU. This study was halted due to “registration process difficulties”. A total of 82 patients were randomized into the Apligraf^® group + SOC (n=33) and SOC alone (n=39). At 12 weeks, wound closure in the Apligraf^® group (n=33) was 51.5% as compared to 26.3% in the SOC group (n=38). The Apligraf^® group achieved complete wound healing over a shorter duration as compared to the SOC group (p=0.059, Log-rank test). The Apligraf^® group took a median of 84 days to heal compared to no median reported in the SOC group due to less than 50% achieving wound closure. An average of 1.8 Apligraf^® applications over 12 weeks were utilized. Limitations include no median time to heal in the SOC group, halting of the study, and lack of blinding.⁴⁷

Additional evidence of Apligraf^® is reviewed in the following sections AmnioBand^®,⁴¹ Epifix^®,⁴⁸ TheraSkin^®, the retrospective study (n=226) by Kirsner et al.⁴⁹ and in the VLU section. In the Kirsner et al. study, the average number of applications over 4 weeks was 3.5 for EpiFix^® and 2.5 for Apligraf^®.⁴⁹

Artacent^®

Sledge et al.⁵⁰ performed an observational study which included 26 patients with DFU (4.65±4.89cm²) with failure to heal by >50% after 2 to 4 weeks of SOC treatment and randomized to a larger clinical trial that had been discontinued for logistical reasons. Patients were randomized to weekly or biweekly applications of dual layer amniotic membrane plus SOC for 12 weeks. A total of 17/26 (65%) achieved complete closure. The small sample size precluded meaningful comparison between the weekly and biweekly applications. Limitations of the study include risk of bias, observational design, lack of control group, variability in length of SOC treatment, small sample size, and inability to determine if healing was impacted by the product, as well as frequency of product applications or other factors. The evidence was not sufficient to determine if the product was effective for treatment of DFU.

Biovance^®

An observational study included 179 chronic wounds of which 47 were DFU. Twenty-eight ulcers studied had failed 32 previous treatments with 1 or more advanced biologic treatments and 48.4% of these showed improvement after treatment with Biovance^® within an average of 8 weeks. For all wound types (n=166) the closure rate was 41.6% within 8 weeks with mean application of 2.12 products.⁵¹ The study was limited by not reporting wound reduction size, outcomes for wound types were not reported separately and small sample size, lack of randomization, blinding or controls. Without a control group, the percentage of wounds that would have healed with SOC is unknown. Additional evidence includes a case series with 14 subjects.⁵² Evidence was not sufficient to determine the efficacy of this product for wound healing.

DermACELL^®

A multi-centered, RCT compared healing rates with a human ADM (DermACELL^®) (n=53), SOC (n=56) and a second ADM (Graftjacket^™) (n=23) for full thickness DFU. One to 2 applications of the graft were applied at the discretion of the investigator for 16 weeks. The DermACELL^® arm had a significantly higher proportion of completely healed ulcers than the SOC arm (67.9% vs 48.1%; p=0.0385) and a nonsignificant higher proportion than the Graftjacket^™ arm (67.9% vs 47.8%; p= 0.1149). There were no serious AEs related to the graft reported.⁵³ This same study population was reported by Cazzell et al. after subjects were followed for 24 weeks.⁵⁴ These 2 studies were published in 2 different journals but share authors and data sets are identical, so it appears to be the same study population. The DermACELL^® group had a significantly higher healing rate over SOC at 16 and 24 weeks which was not found in the Graftjacket^™ group. Closed ulcers in the single application DermACELL^® arm remained healed at a significantly greater rate than the conventional care arm at 4 weeks post termination (100% vs 86.7%; p =0.0435). Strength of the studies include randomization, consistent diagnostic criteria with the same type of ulcers and 24-week follow-up.⁵⁴ Limitations of the study include variation in SOC, lack of blinding, short-term follow-up, randomization methodology was not reported, and some concerns of risk of bias.

A prospective single arm, multicentered trial included 61 participants with large and complex DFU, with the average size 29.0 cm² and, 59/61 had exposed bone. Participants received treatment with ADM allograft (DermACELL^®). Up to 1 additional application was allowed if the wound required further coverage for exposed deep tissue, was less than 75% granulated at 4 weeks or less than 50% granulated after 8 weeks. Wound measurements were validated with a laser measurement device. Fourteen participants did not complete the 16 weeks of which 8 required surgical intervention for their targeted wound, but there were no AEs related to the allograft. The authors report 100% granulation and 31.9% closure by 16 weeks in the per protocol group with 9 receiving a second application with an average of 1.2 applications. In the intention to treat group 90.2% achieved granulation and 24.6% closure by 16 weeks with an average of 1.2 applications. The study did not extend past 16 weeks, but it was postulated that many of the wounds not healed would continue healing if allowed additional time. This underscores the challenges in this difficult population of large ulcers extending to bone.⁵⁵ This study is limited by lack of control arm or randomization, short term follow up, and small sample size.

Additional studies are in the form of case reports and series. Breast reconstruction and burns were not reviewed in this analysis.

Dermagraft^®

A RCT study at 35 centers enrolled 314 patients and reported on 245 with chronic DFU. Patients meeting the inclusion criteria were matched and randomized to Dermagraft^® or SOC. Subjects received up to 7 additional applications at weekly intervals over the course of the study. The authors reported complete wound closure in 30% (39/130) in the Dermagraft^® group as compared to 18.3% (21/115) in the SOC group at week 12. They reported similar AEs in both groups with fewer ulcer related AEs in the Dermagraft^® group.⁵⁶ The study is limited by concerns for risk of bias, and short term follow-up.

In 1996 a multi-centered RCT with 50 subjects was conducted comparing Dermagraft^® at 3 different doses to SOC for DFUs over a 12-week period. Treatment groups included weekly application of 1 piece of Dermagraft^® for a total of 8 applications (Group A [n=12]), application of 2 pieces of Dermagraft^® every 2 weeks for a total of 8 (Group B [n=14]), application of 1 piece of graft every 2 weeks for a total of 4 (Group C [n=11]), and SOC alone (Group D [n=13]). The authors noted that Group A demonstrated statistically significant wound healing (p=0.017) by 12 weeks with a 50% closure rate compared to 50%, 18.2% and 23.1% closure rates for groups B, C, and D respectively.⁵⁷ This study is limited by small sample size, short-term follow-up, lack of blinding, and risk of bias.

Dermagraft^® was reported in a RCT comparing Dermagraft^® to TheraSkin^® for DFU⁵⁸ (see TheraSkin^® section). Dermagraft^® is also reviewed in the VLU section.

EpiCord^®

Tettelbach et al.⁵⁹ performed a multi-center, RCT, to compare dehydrated human umbilical cord (i.e., EpiCord^®) with SOC to treat chronic DFU. A total of 155 patients were treated and included in the ITT analysis: 101 in the EpiCord^® group and 54 in the SOC group. The healing rate at 12 weeks was 70% (71/101) for the EpiCord^® group and 48% (26/54) in the SOC group (p=0.0089). The median number of EpiCord^® allografts applied was 7 (range 2-12). Strengths of this study include a control group (alginate), larger sample size and low risk of bias. Limitations of the study include lack of blinding, and short-term follow-up.

EpiCord^® was also included in a systematic review in which the authors conclude biological skin substitutes were 1.67 times more likely to heal by 12 weeks than SOC dressings (p<0.00001). They also state that further studies are needed to determine the benefits of the different products and the long-term implications of these products.⁶⁰

EpiFix^®

A RCT aimed to investigate wound healing for DFU with EpiFix^® compared to SOC. Twenty-five subjects were randomized to EpiFix^® with replacement of the product every 2 weeks or SOC and followed for 6 weeks. The authors report wound healing in 92% of the EpiFix^® group and 8% of the SOC group. “The EpiFix^® material, placed on an every other week regimen, aggressively closed the wounds under consideration in a far shorter time than standard wound treatment.”⁶¹ Sample size was too small to draw conclusions based upon these results and the study was challenged by lack of blinding and high risk of bias. The outcomes in the SOC arm were concerning because the results were well below those reported by other studies for SOC treatment. In addition, the protocol SOC was not defined in the paper.

A 2016 multi-center RCT with 100 participants compared dehydrated human/amnion/chorion membrane (dHACM) (i.e. EpiFix^®) to SOC and bioengineered skin substitute (Apligraf^®), concluding that dHACM was superior in achieving complete wound closure within 4–6 weeks. The proportion of wounds achieving complete closure within the 12-week study period were 73% (24/33), 97% (31/32), and 51% (18/35) for Apligraf^®, EpiFix^® and SOC, respectively (adjusted p=0.00019). Mean time-to-heal was 47.9 days (95% CI: 38.2–57.7) with Apligraf^®, 23.6 days (95% CI: 17.0–30.2) with EpiFix^® group and 57.4 days (95%CI: 48.2–66.6) with the SOC only group (adjusted p=3.2x10.7). Median number of grafts used per healed wound were 6 (range 1–13) and 2.5 (range 1–12) for the Apligraf^® and EpiFix^® groups. The study was limited by small sample size, lack of blinding and high risk of bias.⁴⁸

A multi-centered RCT which included 110 patients with DFU was undertaken to determine whether EpiFix^® led to improved wound healing compared to SOC. Both ITT and per-protocol participants receiving weekly EpiFix^® (n=47) were significantly more likely to completely heal than those not receiving EpiFix^® (n=51), ITT was 70% vs 50%, p=0.0338, per-protocol was 81% vs 55%, p = 0.0093).⁶² The study had a low risk of bias. Limitations included the short term follow up, and lack of blinding.

EpiFix^® was included in multiple systematic reviews and meta-analysis:The AHRQ report⁴, Cochrane Systematic Review³⁰ and Paggiaro systematic review.⁶³ There is also a National Institute for Health and Care Excellence (NICE) innovation briefing on EpiFix^®.⁶⁴ Additional literature includes case series, lab studies and additional studies in the VLU population reviewed in that section below.

A prospective RCT was performed to compare weekly applications of Apligraf^® (n=20), EpiFix^® (n=20), or SOC (n=20) effectiveness in DFU. Three sites and 65 subjects entered the 2-week run-in period while 60 were randomized to each treatment group. Wound closure was as follows for 4 and 6 week follow up: EpiFix^® (85% and 95%), Apligraf^® (35% and 45%), and SOC (30% and 35%). The mean number of applications used in the Apligraf^® group was 6.2 per patient and 2.15 for EpiFix^® in 6 weeks. All EpiFix^® patients exited the study by the 6 week follow up while 20% of the Apligraf^® patients remained unhealed at 12 weeks. Limitations include that the study was inadequately powered to reach statistical significance between Apligraf^® and SOC group at 6 weeks, short duration of follow up after patient healing period, and the lack of comparison of 12-week healing rates due to missing outcome data which created a high risk of bias.⁶⁵

Grafix^®

Lavery et al.⁶⁶ performed an RCT to contrast the effectiveness of a human viable wound matrix (hVWM) (i.e., Grafix^®) to SOC for ulcer closure in chronic DFU. Patients in the active treatment group received SOC plus an application of Grafix^® once a week (± 3 days) for up to 84 days (blinded treatment phase) and the control group received SOC ulcer therapy once a week (± 3 days) for up to 84 days. The percentage of patients who attained complete ulcer closure was substantially higher in the active treatment group (62%) compared with the control group (21%, p=0.0001). The median time for healing was 42 days in the active treatment arm contrasted with 69.5 days in the control arm (p=0.019). There were less AEs in the active arm (44% vs 66%, p=0.031) and less ulcer-related infections (18% vs 36.2%, p=0.044). The authors concluded that treatment with Grafix^® substantially improved DFU healing in comparison to SOC therapy. Limitations of the study included lack of blinding, short-term follow-up, and high risk of bias.

Additional literature includes a retrospective report of 441 wounds from a healthcare database to evaluate the proportion of DFU that achieved complete closure with viable cryopreserved placental membranes (vCPM) (Grafix^® PRIME and Grafix^® CORE) as compared to standard wound care by 12 weeks and the number of wound related infections and amputations. They reported closure in 59.4% of 350 wounds with the median treatment duration of 42 days and a median of 4 applications (95% CI 4-5) of vCPM with a 3% rate of amputation and an incidence of 2% for infections. Smaller wounds were quicker to heal. There was no comparison to wounds that did not have vCPM applied. Limitations of the study include its retrospective nature, lack of standardized treatment practices, no comparator group, lack of a control cohort, risk of incomplete records, and variabilities in evaluations.⁶⁷

Grafix^® CORE

Frykberg et al.⁶⁸ conducted a prospective, multicentered, open labeled single arm RCT using vCPM (Grafix^® CORE, Osiris Therapeutics, Inc) in 31 complex DFU with exposed deep structures. The wounds were cleaned and debrided weekly with weekly application of vCPM and protective foam dressings. Fifty-nine percent achieved complete wound closure by 16 weeks. These data show that vCPM is a safe and effective option for the successful management of complex wounds with exposed tendon and bone. As vCPM was not combined with other advanced modalities (i.e., NPWT) during the course of treatment in this study, it would be of interest in the future to investigate the cumulative benefits of vCPM as part of a multimodal approach to complex wounds with exposure of deep structures or bone. This study was limited by a lack of comparison to standard wound care, no disclosure of funding sources suggesting higher potential risk of bias and high dropout rate given the small number of patients enrolled. Evidence was not sufficient to determine the efficacy of this product for wound healing.

Graftjacket^™

A pilot study was conducted to evaluate the potential role of Graftjacket^™ in ulcer management with 40 subjects comparing Graftjacket^™ to gauze dressings with a suggested potential role in ulcer management. Rates of healing were reported as decrease in wound area by 67.4% in the Graftjacket^™ group compared to 34% in the SOC group at 4 weeks.⁶⁹ A second RCT study was conducted to evaluate the effectiveness of Graftjacket^™ for chronic non-healing lower extremity wounds. Subjects received a single application of Graftjacket^® (n=14) compared to controls treated with gauze dressings (n=14) and followed for 16 weeks. A total of 85.71% of the treatment group ulcers were healed compared to 28.57% of the control group at the conclusion of the study (p=0.006).⁷⁰ Limitations of both studies included a small sample size and high risk of bias.

A multicentered RCT compared subjects with DFU receiving acellular matrix (Graftjacket^™ Regenerative Tissue Matrix) (n=47) to SOC (n=39). The authors reported a complete healing time of 69.6% at 5.7 weeks for the treatment group compared to 46.2% at 6.8 weeks for the control group. The proportion of healed ulcers between the groups was statistically significant (p= 0.0289) with odds of healing 2.7 times higher in the study group than the SOC group. Subjects received a single application and were followed to 12 weeks. Six AEs were reported but not related to the graft except in 1 case where the graft was no longer on the wound.⁷¹ Strengths of the study include randomization and defined control group with certain limitations noted such as a short term follow up and high risk of bias.

These 3 studies were pooled in a meta-analysis (n=154) comparing Graftjacket^™ to SOC and reported a statistically significant reduction in ulcer size in 1.7 weeks and a fourfold improvement in the chance of healing in the Graftjacket^™ group. The authors conclude that a single application of this product after sharp debridement and offloading may improve healing for DFU and the model used predicted an average of 1.7 weeks reduction in healing time with this approach. The median number of applications per patient, after initial application, was 1 (range 1-15). There were differences in outcome measures in the 3 studies challenging the pooled results. Limitations include high risk of bias including publication and reporting biases, study selection biases, incomplete data selection, and a high risk of bias, due to small sample sizes and differences in endpoints. ⁷²

Additional studies include 2 RCTs in which Graftjacket^™ was compared to DermACELL^® and SOC, but with only 23 subjects in Graftjacket^™ arm, the study was not sufficiently powered to draw conclusions.⁵⁴ Other investigations (see section on DermACELL^®), include a Cochrane review analysis²⁸ and multiple studies investigating the role of the product in tendon repair and breast reconstruction.

Integra^®

Driver et al. conducted The Foot Ulcer New Dermal Replacement Study (FOUNDER), a RCT with 153 patients in the control arm who received SOC treatment and 154 patients in the active treatment arm received Integra^® Dermal Regeneration Matrix for DFU. Both groups underwent 14-day run-in periods where they received SOC treatment and eligible patients were randomized with software algorithm and ulcers were measured at onset. Complete closure of the ulcer at 16 weeks was significantly greater in the active group (51%; 79/154) in comparison to the control group (32%; 49/153, p=0.001). There were no significant AEs in either group.⁷³ Strength of the study included the randomized design, large sample size, control group, multi-centered, run-in period, set wound type and inclusions/exclusion criteria. Limitations of the study include lack of double blinding, the short-term follow-up and high risk of bias.

A prospective pilot study evaluated 10 patients treated with Integra^® bilayer wound matrix for DFU. The authors report 70% (7/10) achieved complete wound healing by 12 weeks.⁷⁴ This study is limited by study design, very small sample size and short-term follow-up. Additional literature includes case reports, series and retrospective reviews.

Kerecis^® Omega3

A double-blinded RCT compared fish skin allograft (Kerecis^® Omega3 Wound) to dehydrated human amnion/chorion membrane allograft (EpiFix^®) for induced wounds. Subjects (n=170) received punch biopsies and the graft was placed over the induced wound. The subject and assessor were blinded to the treatment group. Wounds treated with fish skin healed significantly faster (hazard ratio 2.37; 95% CI: (1.75–3.21; p = 0.0014) compared with wounds treated with EpiFix^® over a 28-day period. The average was 1.6 applications per subject for the Kerecis^®Omega3 wound and 1.4 applications for Epifix^®.⁷⁵ This was a high-quality study, but the results were not applicable to chronic non-healing wounds.

A multi-centered RCT compared fish skin allograft (Kerecis^® Omega Wound) + SOC to SOC alone in 49 patients with chronic DFU after a 2-week screening period. At 12 weeks, 16 of 24 patients' DFU (67%) in the fish skin arm were completely closed, compared with 8 of 25 patients' DFU (32%) in the SOC arm (p=0.0152 [N=49]; significant at p<0 .047). The median number of applications to achieve closure was 5 (in arm 1).⁷⁶ Limitations include high risk of bias due to missing outcome date, small sample size and short-term follow-up period.

Additional literature includes review papers. Seth et al. summarizes case reports, case series and retrospective studies, and noted that additional RCTs are ongoing.^77,78 This evidence is insufficient to validate net positive outcomes for DFU or VLU.

MatriStem^®

A multi-centered observational study was conducted at 13 U.S. centers and included 56 subjects comparing MatriStem^® MicroMatrix (MSMM) and MatriStem^® Wound Matrix (MSWM) (porcine-derived) (n=27) to ulcers treated with Dermagraft^® (n=29) for DFU. The matrix was applied weekly until wound closure or 1 application per week without wound closure whichever came first to a maximum of 8 applications. Subjects were followed for 6 months for ulcer recurrence with one recurrence in both groups. There were no statistically significant differences between the 2 groups in the following: complete wound closure at day 56 (p=0.244), change in wound size over 8-week treatment period (p=0.762); complete wound closure at day 70 (p=0.768); or mean time to closure (p=0.523).⁸ This study's strength includes the multicentered sites and following subjects for 6 months for recurrence, but only 10 subjects were followed for this duration. The small sample size is not sufficient to determine efficacy of this product for wound healing.

Microlyte^® Matrix

Manning et al.⁷⁹ performed an open-label, prospective pilot study to evaluate a bioresorbable polymeric matrix infused with ionic and metallic silver (i.e., Microlyte^® Matrix) as a primary wound contact dressing in the treatment of 32 patients (median age of 62 years) with a total of 35 hard-to-heal wounds along with SOC. The wounds encompassed venous stasis ulcers, DFU, postoperative surgical wounds, burn wounds, and chronic, non-pressure lower extremity ulcers unresponsive to standard protocols of care. Of the 35 chronic wounds, the majority consisted of venous stasis ulcers (54%) (19/35), followed by DFU (23%; 8/35). The mean wound surface area at the start of the study was 6.7 cm² (range 0.1 cm² – 33 cm²); the median wound surface area was 2.1 cm². These wounds were considered as nonhealing for a median of 39 weeks (range, 3-137 weeks) and suspected to have persistent microbial colonization that had not responded to standard antimicrobial products and antibiotics.

The micrometer-thick bioresorbable matrix conforms closely to the underlying wound bed to exert localized and sustained antimicrobial action of noncytotoxic levels of silver. The matrix was applied to the wounds once every 3 days to provide a scaffold for uniform loading of silver nanoparticles and a template for cells migration and then covered with a secondary dressing. Any residual matrix still in the wound was not removed due to the bioresorbable nature of the matrix. Three patients were lost to follow-up after initial application. At 3 weeks, 72% of wounds (22/32) had an average wound area reduction of 66%. Of the 16 venous stasis ulcers, 11 improved by an average healing rate of 60%, and 6 of 8 DFU improved by an average wound area reduction of 79%. At the 3-week assessment, the burn wound, and postoperative wounds had an average wound area reduction of 38% and 58%, respectively. By 12 weeks, 91% of wounds (29/32) either healed completely (i.e., fully re-epithelialized) or improved substantially with an average wound area reduction of 73%. The venous stasis ulcers and DFU had an average wound area reduction greater than 75%, with visual signs of healthy granulation tissue formation and re-epithelialization. The study had certain limitations which included a small sample size, and use of the same clinical investigator who performed all assessments during the study.⁷⁹ There was not sufficient evidence to determine the efficacy of this product for wound healing.

Mirragen^®

There has been interest in bioactive glass as a pathway to wound healing due to postulated ability to release ions that can stimulate processes, such as hemostasis, antibacterial efficacy, epithelial cell migration, angiogenesis, and fibroblastic cell proliferation.⁸⁰ A literature review of bioactive glass applications introduced the potential of this product and called for further research to understand the clinical role.^81,82 A randomized trial was conducted to evaluate a unique resorbable glass microfiber matrix (Mirragen^®; Advanced Wound Matrix) compared to SOC for 12 weeks. All patients received standard diabetic wound care and 20 were treated with the matrix while the others received SOC only. The primary endpoint was non-infected wound healing at 12 weeks. The authors report that in the intent-to-treat analysis results at 12 weeks showed that 70% (14/20) of the Mirragen^®-treated DFU healed compared with 25% (5/20) treated with SOC alone (adjusted p=0.006).⁸³ Strengths of the study include robust design, randomization, ITT analysis, and multiple sites. While the study was adequately powered per sample sized calculation the large drop-out in the SOC group (12/20) resulted in high-risk bias due to missing outcome data. Combined with the small sample size, lack of blinding and short-term follow-up ranging from 6-12 weeks there was not sufficient evidence to understand safety, effectiveness, and long-term outcomes of this product.

NEOX^® CORD/TTAX01

A multi-centered prospective trial of cryopreserved human umbilical cord (TTAX01; NEOX^®) enrolled 32 subjects with complex wounds which extended to muscle, fascia or bone with underlying osteomyelitis with a mean duration of 6.1 ± 9.0 (range: 0.2–47.1) months and wound area at screening of 3.8 ± 2.9 (range: 1.0–9.6) cm² which was increased to 7.4 ± 5.8 (range: 1.1–28.6) cm² after aggressive debridement. Initial closure occurred in 18 of 32 (56%) wounds, with 16 (50%) of these having confirmed closure in 16 weeks with a median of 1-product application. Ulcers with biopsy confirmed osteomyelitis (n=20) showed initial closure in 12 (60%) and confirmed closure in 10 (50%). Mean healing time was 12.8 ± 4.3 weeks. The average number of applications was 1.5 ±0.8 applications (median of 1, range 1–3) over 16 weeks.⁸⁴ These same patients were reported on in a follow-up report that included 30 subjects with evaluation for safety, while subjects with a remaining open or closed index wound (n=29) were evaluated for efficacy. One subject had his unhealed wound removed in a minor amputation in the previous study. They were followed for 1 year and the AEs reported were all typical for the population under study, and none were attributable to NEOX^®. One previously healed wound re-opened, 1 previously unconfirmed closed wound remained healed, and 9 new wound closures occurred, with 25 of 29 (86.2%) healed in the ITT population. This included use of additional products, minor amputation (n=2) and 1 major amputation.⁸⁵ Limitations include small sample size, lack of controls, and no randomization. However, this investigation did assess complex wounds that are rarely included in clinical studies. Additional literature includes a basic science report, case series and small retrospective reports. These studies have inherent limitations due to the small sample size and observational design and there is no way to be certain that the treated wounds would have similar healing as compared to other skin substitutes or SOC. The potential benefit in a complex population (exposed tendon, muscle, and bone) warrants further investigation.

NeoPatch™

A multi-centered prospective study was conducted with 63 patients with chronic DFU. Wounds were classified by size into ‘small’ (≤2.0 cm²), ‘medium’ (>2.0–4.0 cm²), and ‘large’ (>4.0–25.0 cm²). After a 2-week run in period, patients were treated with chorioamniotic allograft (NeoPatch™) on a weekly basis until the study period ended or wound closure to a maximum of 11 applications. At week 12, 13/23 small ulcers, 5/15 medium, and 1/10 large ulcers achieved closure, with a mean number of applications of 6.2, 6.6, and 8.0, respectively. The mean for the entire group was 40% closure (19/48) with 6.4 applications in 12 weeks. Of the AEs reported, most were related to the ulcer with no reported AEs attributable to the allograft.⁸⁶ Limitations of the study include the lack of randomization, control group, short term follow-up, small sample size, and potential risk of bias.

OASIS^® ULTRA Tri-Layer Matrix

A RCT comprised of 11 centers and 82 subjects with DFU was completed to compare clinical outcomes of patients treated with tri-layer OASIS^® vs SOC. Patients were randomized into OASIS^® group (n=41) or SOC (n=41) group and evaluated for 12 weeks or until complete wound closure was achieved. The OASIS^® group achieved a significantly greater number of complete closures compared to the SOC group (54% vs 32%, P=0.021) at 12 weeks. Limitations include unblinded design, short duration of follow up, and high risk of bias. Strengths of the study were comprised of the randomization process and use of a digital wound measurement device.⁸⁷

PriMatrix^®

Lantis et al. (2021) conducted a multicenter RCT to evaluate the safety and efficacy of a fetal bovine ADM (PriMatrix^®) plus SOC vs SOC alone for treating hard-to-heal DFU. Participants (n=226) and 161 completed the protocol with 59.5% (47/79) with wound closure in the PriMatrix^® group and 35.4% (29/82) in the SOC group (p=0.002) in the per protocol analysis. Of wounds that healed, median time to close was 43 days for PriMatrix^® group and 57 days for SOC group. The median number of applications of PriMatrix^® to achieve closure was 1.⁸⁸ AEs were similar between groups and no product-related serious AEs occurred. The author noted study limitations such as short term follow up, inability to blind investigators or subjects to treatment type, patient selection bias towards healthier patients, and an overall high risk of bias.

A prospective trial reported on 55 subjects from 9 centers with DFU treated with PriMatrix^® and followed for 12 weeks. 76% healed by 12 weeks with a mean time to healing of 53.1 ± 21.9 days. The mean number of applications for these healed wounds over 12 weeks was 2.0 ± 1.4, with 59.1% healing with a single application of PriMatrix^® and 22.9% healing with 2 applications. For subjects not healed by 12 weeks, the average wound area reduction was 71.4%.⁸⁹ Study is limited by observational design without control group.

Additional literature includes a basic science report⁹⁰ and a retrospective review.^91,92

PuraPly^®

A prospective, noninterventional, multicentered study was conducted to evaluate the effectiveness of purified native type I collagen matrix plus polyhexamethylene biguanide antimicrobial (PHMB) on cutaneous wounds (PuraPly^® AM). A cohort of 307 patients with VLU (n=67), DFU (n=62), pressure ulcers (n=45), post-surgical wounds (n=54), and other wounds (n=79) were treated with PuraPly^® and followed for 12 weeks. The number of applications ranged from 1-2 (21.8%) to 10 (<2%). They report that 73.2% of wounds were reduced from baseline and 63.4% had reached ≥70% reduction in area at 12 weeks with 37% of wounds achieving complete wound closure at 12 weeks. The average number of applications was 5.2 with 21.8% receiving 1 or 2 applications (21.8%) and <2% receiving 10 or more applications. No AEs were reported related to the product.⁹³ The study is limited by lack of a control group, blinding or randomization, short term follow-up, and high-risk of bias. While this study shows promising results, it is difficult to determine whether the treated wounds would have similar healing as compared to other skin substitutes or SOC. Additional literature includes a case series and retrospective review.

Restrata^®

In a 2017 report, Restrata^® is introduced as a fully synthetic, resorbable electrospun material (Restrata^® Wound Matrix) that exhibits structural similarities to the native ECM. The product was tested in a swine model.⁹⁴ A retrospective review of the product reported on 82 ulcers in patients with DFU (n=34) or VLU (n=34) and other wounds (n=14). They report 85% of the wounds achieved complete closure at 12 weeks.⁹⁵ Limitations include study design without controls not sufficient to conclude if the outcome were directly related to the novel product and insufficient follow-up time to establish safety.

TheraSkin^®

A RCT trial investigated 50 subjects with DFU were treated with cryopreserved bioactive split thickness skin allograft (TheraSkin^®) and 50 were treated with SOC (collagen alginate dressing). The authors reported at 12 weeks 76% (38/50) of the TheraSkin^® group vs 36% (18/50) for the SOC group achieved healing. The number of allografts to achieve healing was not reported.⁹⁶ Strengths of the study include randomization, ITT analysis, and low risk of bias. Despite the high dropout rate in SOC arm (n=19) the investigator used the last observation carried forward method to account for missing outcome data in the SOC group. Limitations in the study include small sample size, lack of blinding, and short-term follow-up.

A prospective study reported on 17 patients with DFU treated with the bioengineered skin substitute (Apligraf^®) and 12 were treated with a cryopreserved split thickness skin allograft (TheraSkin^®). Most received a single application with the decision to reapply left to the treating provider. The authors report 41.3% of the ulcers treated with Apligraf^® and 66.7% of the ulcers treated with TheraSkin^® were closed at 12 weeks, 47.1% treated with Apligraf^® closed at 20 weeks. The number of closed TheraSkin^® treated ulcers remained 66.7% at 20 weeks. The average number of applications of Apligraf^® was 1.53(SD=1.65). The number of applications of TheraSkin^® was 1.38 (SD=0.29). There were no significant AEs reported.⁹⁷ Limitations of the study include small sample size, lack of control, short term follow up, and high risk of bias.

Sanders et al.⁵⁸ performed a multi-centered RCT to contrast an in vitro-engineered, human fibroblast-derived dermal skin substitute (HFDS) (i.e., Dermagraft^® to a biologically active cryopreserved HSA (i.e., TheraSkin^®) in the treatment of DFU. The primary objectives were to establish the relative number of DFU healed (100% epithelization without drainage) and the number of grafts needed by week 12. Twenty-three eligible patients were randomly assigned to the Dermagraft^® treatment group (12 patients) (mean age 57) or the Theraskin^® treatment group (11 patients) (mean age 60). Patients in the TheraSkin^® group received a product application every other week and patients in the HFDS group were treated every week with SOC. After the week 12 visit, no additional biologically active products were used in either treatment group. Patients with incomplete ulcer closure continued to be evaluated through week 20; subsequent treatment was then provided outside the scope of the study. At week 12, 7 (63.6%) ulcers in the TheraSkin^® treatment group vs 4 (33.3%) in the Dermagraft^® treatment group were healed (P=0.0498). At the end of week 20, 90.91% of ulcers in the Theraskin^® group vs 66.67% of ulcers in the Dermagraft^® group were healed (P=0.4282). The average of 8.92 applications (range 6-12 applications) in up to 20 weeks for Dermagraft^® and mean applications of 4.36 (range 2-7) in up to 20 weeks for Theraskin^®.⁵⁸

Time to healing in the TheraSkin^® group was substantially less (8.9 weeks) than in the HFDS group (12.5 weeks) (log-rank test, P=0.0323). The results of this study showed that, after 12 weeks of care, DFU treated with HSA were probably twice as likely to heal as DFU managed with Dermagraft^® with about half the number of grafts required. Limitations noted for this study include small sample size, short-term follow-up and high risk of bias.⁵⁸

Additional literature includes 2 large retrospective matched cohort studies,^98,99 several cost-analysis, and an animal model study.

Clinical Trials for Skin substitute graft/CPTs for VLU

AmnioBand^®

Serena et al.¹⁰⁰ performed an open-label, multicenter RCT comparing 2 application treatments of dehydrated human amniotic and chorion allograft (dHACA) (i.e., AmnioBand^®) with SOC vs SOC alone for the treatment of 60 patients with VLU. Patients were randomized into 1 of 3 study groups: SOC alone (control), weekly AmnioBand^® with SOC or biweekly AmnioBand^® with SOC (20 patients per group). At 12 weeks, healing rates were 30/40 (75%) in the 2 AmnioBand^® groups and 6/20 (30%) in the SOC group; p=0.001. Treatment with AmnioBand^® continued to be significant after adjustment for wound area (p=0.002), with an odds ratio of 8.7 (95% CI: 2.2-33.6). Only 6 VLU (30%) were healed in the SOC group contrasted to 15 (75%) in the weekly AmnioBand^® group (p= 0.02) and 15 (75%) in the biweekly AmnioBand^® group (p= 0.02). There were no significant differences in the proportion of wounds with percent area reduction (PAR) ≥40% at 4 weeks among all groups. All analyses used the ITT approach, and the risk of bias was low. Limitations include lack of blinding and short-term follow.¹⁰⁰

Apligraf^® (formerly Graftskin^®)

Falanga et al. (1998) performed a multicenter RCT to evaluate a allogeneic human skin equivalent (HSE) Apligraf^® group (n=146) vs SOC (n=129) in 275 patients with VLU. At 6 months, 63% Apligraf^® vs 49% SOC patients were healed. Median time to complete wound closure was 61 days in the Apligraf^® group vs 181 days in the SOC group. An average of 3.34 applications of Apligraf^® per patient were utilized.¹⁰¹ There were some concerns for risk of bias due to per protocol analysis only as well as short-term follow-up.

A prospective RCT included 120 patients with hard to heal VLU for a duration of greater than 1 year. Patients were randomized into an Apligraf^® plus compression therapy (n=74) or standard compression therapy (n=48) groups. Wound closure at 6 months was reported as 47% for the Apligraf^® group vs 19% for the control group. The authors conclude at 6 months, that patients treated with Apligraf^® were twice as likely to achieve complete wound closure as compared to standard compression therapy. They report Apligraf^® was over 60% more effective than the control in achieving wound closure. Limitations include high risk of bias, and lack of blinding.¹⁰²

A prospective randomized pilot study was conducted to estimate the relative difference in the effectiveness of Apligraf^® and Theraskin^® and compression therapy for the treatment of VLU. A total of 31 participants were randomized and they reported a higher healing rate in the Theraskin^® cohort (93.3%) as compared to the Apligraf^® cohort (75.0%) at 12 weeks, but it was not statistically significant. At 20 weeks follow up, the Theraskin^® cohort remained at a 93.3% vs Apligraf^® cohort at an 83.3% healing rate. The mean number of applications was 3.33 in the Apligraf^® group and 2.27 in the Theraskin^® group for 12 weeks. Limitations of this study include low sample size, and high risk of bias. There were no AEs reported.¹⁰³

DermACELL^®

Cazzell¹⁰⁴ conducted a multicenter, RCT designed to evaluate the safety and efficacy of human decellularized acellular dermal matrices (D-ADM) (DermACell^® AWM) (n=18) contrasted with SOC (n=10) in patients with chronic VLU. The study participants were randomly assigned to the D-ADM (i.e., DermACELL^® AWM) treatment arm or a SOC treatment arm in a 2:1 ratio. A blinded, independent adjudicator also assessed the healing condition of all ulcers. Patients could have a maximum of 2 DermACELL^® applications, which included the first application at baseline and 9 (50%) received a second application during the study. At 24 weeks, patients in the DermACELL^® arm demonstrated a strong trend of reduction in the ulcer area, with a mean reduction of 59.6%, in comparison to the SOC arm, with a mean reduction of 8.1%. Also, the ulcer areas in the SOC arm increased more than 100% in size for one-third (3/9) of the patients. Furthermore, healed ulcers in the DermACELL^® arm stayed closed at a significantly greater rate after initial confirmation of complete ulcer closure than healed ulcers in the control arm. Limitations noted for this study included a small patient population with an unbalanced proportion between the 2 groups (2:1) that ensured a low probability of achieving statistical significance, insufficient criteria for investigators to follow when deciding if a second application would be appropriate, and a short-term follow-up. The overall risk of bias was low.

Dermagraft^®

Harding et al.¹⁰⁵ conducted a multicenter RCT that assessed the human fibroblast-derived dermal substitute (HFDS) (i.e., Dermagraft^®) plus compression therapy contrasted with compression therapy alone in the treatment of VLU. The primary outcome variable was the proportion of patients with completely healed study ulcers by 12 weeks. Sixty-four (34%) of 186 patients in the Dermagraft^® group demonstrated healing by week 12 compared with 56 (31%) of 180 patients in the control group (P=0.235). For ulcers ≤ 12 months duration, 49 (52%) of 94 patients in the Dermagraft^® group contrasted with 36 (37%) of 97 patients in the control group healed at 12 weeks (P=0.029). For ulcers ≤ 10 cm², complete healing at week 12 was shown in 55 (47%) of 117 patients in the HFDS group contrasted with 47 (39%) of 120 patients in the control group (p=0.223). The most common AEs were ulcer infection, cellulitis, and skin ulcer. The occurrence of AEs was not significantly different between the treatment and control groups. Statistical significance was not achieved for the primary outcome of complete closure in patients with VLU completely healed by 12 weeks. The study had some concerns for risk of bias due to high dropout rate and lack of validation of outcome measurements.¹⁰⁵

EpiFix^®

A multi-centered, RCT was conducted to evaluate a dehydrated human amnion/chorion membrane allograft (EpiFix^®) (n=53) with SOC to SOC alone (n=31) for VLU. Subjects randomized to allograft received 1 (n=26) or 2 applications (n=27). At 4 weeks, 62% in the allograft group and 32% in the control group showed a greater than 40% wound closure (p=0.005), and wound size reduction of 48.1% and 19%, respectively. The authors reported the group with 2 applications (baseline and 2 weeks later) had the fastest healing time.¹⁰⁶ Limitations include lack of blinding, small sample size, and short-term follow up.

Another multi-centered RCT comparing EpiFix^® + SOC to SOC alone (multilayer compression therapy) for 109 subjects with VLU and followed for 16 weeks. Participants receiving weekly application of EpiFix^® (n=52) and compression were significantly more likely to experience complete wound healing than those receiving standard wound care and compression (n=57) (60% vs 35% at 12 weeks, P=0.0128, and 71% vs 44% at 16 weeks, p=0.0065).¹⁰⁷ Limitations of the study include lack of blinding, short-term follow-up, and high risk for bias.

OASIS^® Products

OASIS^® Wound Matrix:

Landsman et al. conducted an RCT of 26 subjects with DFU. Subjects were randomized and treated with either Dermagraft^® (n=13), or OASIS^® Wound Matrix (n=13) in conjunction with SOC. Wound dressing was applied for 12 weeks and subjects were followed for 20 weeks. Closure rate for OASIS^® was 76.9% and Dermagraft^® is 84.6% with average closure time of 40.90 ± 32.32 days. No statistically significant difference was reported in closure time between the 2 groups. The average number of applications was 2.54 (±0.78) of Dermagraft^® and 6.46 ± 1.39 of OASIS^® in 12 weeks.¹⁰⁸ Limitations include small sample size, short-term follow-up and some concerns for bias.

Niezgoda et al. conducted a multicenter RCT to compare the rate of healing in DFU patients. Patients were randomized to either the OASIS^® Wound Matrix (n=37) group or Regranex^® Gel (n=36) plus a secondary dressing group. At 12-week follow-up, the OASIS^® group achieved 49% wound healing as compared to 28% in the Regranex^® group. Limitations included small sample size, lack of standardization between centers in debridement techniques, frequency of wound dressing changes, lack of blinding, and some concerns for bias.¹⁰⁹

A 2010 RCT was conducted to compare the OASIS^® Wound Matrix (n=25) to SOC (n=25) in VLU. Investigators assessed the wounds weekly and utilized digital planimetry for wound measurement. At 8 weeks complete wound closure was achieved in 80% (20/25) of OASIS^® Wound Matrix patients as compared to 65% (15/23) in the SOC group (p<0.05). A statistically significant difference was reported for mean ulcer duration. Complete healing was achieved in the treatment group, 5.4 weeks, vs 8.3 weeks in the SOC group, (p=0.02). Granulation tissue was considered in cases where complete wound closure was not achieved by 8 weeks. The granulation of tissue increased from baseline to 8 weeks in the OASIS^® group and was reported as 50% and 65%, respectively, while the control group reported a loss of granulation from a baseline of 50% to a decrease of 38% at 8 weeks. Two subjects withdrew from the control group due to relocation. No adverse effects were reported. Limitations include small sample size, lack of blinding, and some concerns for bias.¹¹⁰

Mostow and colleagues conducted a multicenter RCT comprised of 120 patients with VLU to compare the OASIS^® Wound Matrix plus SOC (n=62) to SOC alone (n=58). Following a 2-week screening period, patients were randomized into 1 of the 2 groups and followed for 12 weeks. A total of 19 patients assigned to the SOC alone group crossed over into the treatment group due to a lack of healing at 6 months. Healing was achieved in 26% (5/19) of these patients after receiving an average of 4 applications of the OASIS^® product. The primary outcome was proportion of healed ulcers at 12 weeks. Although the data was still analyzed, 20% of patients were lost to follow-up (12 in each group). At 12 weeks, the treatment group achieved 55% healing as compared to 34% in the SOC group. Ulcer recurrence did not occur in any of the healed patients in the treatment group over a 6-month period. The average number of applications for VLU was 4 (applied to 5/19 crossover patients). A total of 23 AEs were reported and evenly distributed between the 2 groups. Limitations include lack of blinding, small sample size, short duration of follow-up, limited number of wounds evaluated at 6 months, and high risk of bias.¹¹¹

Additional literature is reviewed in Dermagraft^® section. Literature reviewed but not summarized in this LCD includes a retrospective comparative study in the treatment of VLU.¹¹²

Unspecified OASIS^® Products

O’Donnell and associates conducted a systematic review of RCTs to determine if complex wound coverings impacted wound healing as compared to simple wound dressings. A total of 20 RCTs were included and stratified into 3 classes semi-occlusive/occlusive group (n=8), growth factor group (n=7), and HSE group (n=5). Five of the RCTs (25%) yielded statistical significance for improved proportion of ulcer healing in the treatment group over the control: zinc oxide pastes bandage (79% vs 56%) and Tegasorb™ (59% vs 15%) in the semi-occlusive/occlusive group and perilesional injection of granulocyte-macrophage colony-stimulating factor (57% vs 19%) and porcine collagen derived from small-intestine submucosa (OASIS^®; 55% vs 34%) in the growth factor group. In the sole significant RCT from the HSE group, Apligraf^® (63%) was superior to Tegapore™ (49%).”³ See Apligraf^® section.

A 2019 single blinded RCT comprised of patients with DFU compared 8 weeks of treatment using either Dermagraft^® (n=29) or OASIS^® devices (n=31) (active treatment phase) followed by 4 weeks of SOC (maintenance phase), and SOC (n=29) alone. Each treatment group achieved a statistically significant reduction in wound area from weeks 1 to 28. No differences were reported between groups in complete wound closure by 12 or 28 weeks of treatment. Complete wound closure at 12 weeks was Dermagraft^® (8/17) 47.1%, OASIS^® (14/19) 73.7%, and SOC 57.9% (11/19). Complete wound closure by study conclusion was Dermagraft^® (11/17) 64.7%, OASIS^® (15/19) 78.9%, and SOC 73.7% (14/19). The study was an interim report and did not have enough enrolled to meet the sample size needed, and there was a high risk of bias due to per-protocol analysis only for this interim data. The authors were surprised at the higher healing rates for SOC than what was reported in the U.S. literature and postulated that unintentional bias may have resulted in lower efficacy in the SOC group or favoring SOC in their study.¹¹³ The final results were not identified as being published; therefore, there is a potential risk for publication bias. See the Dermagraft^® section.

Romanelli et al. conducted a RCT to compare OASIS^® (n=27) and Hyaloskin^® (n=27) products in the healing VLU at 16 weeks of treatment. Patients were assessed by complete wound healing, time until dressing change, pain and comfort. A total of 82.6% of OASIS^® ulcers achieved complete wound closure as compared to 46.2% of Hyaloskin^® ulcers. Treatment favored OASIS^® treated ulcers which were statistically significant for time to dressing change (p<0.05), pain (p<0.05) and patient comfort (p<0.01). Four patients were lost to follow-up. No AEs were reported. Limitations include self-reporting bias, small sample size, lack of blinding, and some concerns for bias related to randomization.¹¹⁴

Demling and associates reported an interim analysis of a prospective RCT to examine the effectiveness of OASIS^® products compared to SOC in treating VLU. The primary outcome was wound closure at 12 weeks. At 12 weeks, 84 patients were evaluated in which 71% (32/45) of OASIS^® vs 46% (18/39) SOC patients achieved complete wound healing. Significant improvements in the incidence of healing were reported in the OASIS^® patients vs SOC (p=0.018). Interim results were reported on per-protocol analysis rather than the intention to treat population introducing a high risk of bias.¹¹⁵ The final results were not identified in the literature and do not appear to have been published, which potentiates the risk for publication bias.

Talymed^®

A RCT enrolled 82 patients comparing a poly-N-acetyl glucosamine, nanofiber-derived, technology (Talymed^®) to SOC for VLU. Subjects were randomized to treatment with Talymed^® applied once, every other week, every 3 weeks, or SOC alone and followed for wound healing at 20 weeks. At 20 weeks, the proportion of patients with completely healed VLU was 45.0% (n=9 of 20), 86.4% (n=19 of 22), and 65.0% (n=13 of 20) for groups receiving standard care plus Talymed^® only once, every other week, or every 3 weeks, respectively, vs 45.0% (n=9 of 20) for those receiving standard care alone.¹¹⁶ The biweekly application group showed improvement over the SOC arm (p<0.01). Strengths include randomization, blinded investigator, and presence of a control arm. The investigation had limitations which consisted of a small sample size and high risk of bias due to missing outcome data. While these results were promising, the sample size was too small to determine if the outcomes were related to the product. The authors acknowledge that this was a pilot study and there was a need for a larger study to confirm the findings. Further, 2 of the 3 study arms did not show significant differences from the SOC group.

Additional literature consists of case report¹¹⁷ and was included in a systematic review³³ (see above).

Risk of Bias Assessment

A risk of bias assessment was conducted for all RCTs to evaluate them using the same tool and identify areas of potential concern in study designs. Risk of Bias 2 tool¹¹⁸ (RoB2) was used and is described in the Cochrane handbook¹¹⁹ and utilized in Grades of Recommendation, Assessment, Development, and Evaluation (GRADE). This tool is different than the tool used in the AHRQ report⁴ and the other systematic reviews published prior to 2019 (see the section addressing Systematic Reviews) when the updated tool was published. The revised version requires a judgement about the risk of bias arising from each domain-based on answers to the signaling questions. Judgements are ‘Low’, ‘High’, or can express ‘Some concerns’ and included in the evidence review and Table 1 for each product assessed. The overall result must reflect the highest value assigned to any domain. While almost all included studies were funded by industry, this is not an underlying reason to determine that bias exists using RoB2. This tool requires evaluation of multiple aspects of the trial design and assesses if risk of bias was introduced regardless of funding source.

Table 1: Evidence for Covered Products

Table 2: Evidence for Non-Covered Products

Application changes

A 2021 industry-sponsored study presented a retrospective analysis from the Medicare Limited Data set (2015-2018) comparing lower extremity diabetic ulcers (LEDUs) treatment with advanced treatments (AT) defined as cellular and acellular dermal substitutes, compared to no advanced treatments (NAT). Out of 9,738,760 patients identified with a diagnosis of diabetes, 909,813 had a lower extremity diabetic ulcer (LEDU). Patients treated exclusively with AT or NAT were included in the analysis (i.e., patient treated with another type of advanced treatment were excluded). A set of covariates that included patient demographics, LEDU characteristics, year of episode start, prior treatments, prior visits, and comorbidities was identified. Based on this set, propensity scores were used to create 2 comparable groups with similar distributions of observed covariates. In propensity-matched Group 1, AT patients had fewer minor amputations (p=0.0367), major amputations (p<0.0001), ED visits (p<0.0001), and readmissions (p<0.0001) contrasted with NAT patients (12,676 episodes per cohort). The authors then took a second step in the analysis to attempt to determine the effectiveness of AT following parameter for use contrasted to patients with AT not following parameter for use. They reported patients had fewer minor amputations (p=0.002) in those following parameters for use (1,131 episodes per cohort). They conclude AT with skin substitutes were associated with significant reduction in major and minor amputations, emergency room visits, and hospital readmissions compared to those without AT. They also conclude that following the parameters improved outcomes.¹⁷ The study is limited by lack of blinding and randomization which restricts the ability to determine if these outcomes were directly related to the treatment with skin substitute. It is unclear whether the study considered a number of factors that would be expected to influence outcomes, including visit frequency, compliance with care, infection treatment, and the use of additional products/treatments. It is difficult to draw the conclusion that the improvement was due solely to the AE with skin substitutes and not related to other factors from a retrospective study.

The study does report on frequency of application. Patients who received AT with skin substitute grafts (propensity-matched Group 1) had 3.7 (3.6) applications on average (n=12,313). In Group 2 the average number of applications in the parameter for use group was 4.9 (3.8) (n=1131) and in the not following parameter for use group the average was 3.5 (3.3).¹⁷

The number of applications in the reported literature is variable and differs among products. Factors to consider include whether the application was made per protocol, whether those protocols require a product change or is at the discretion of the provider, and any applications made on an as needed basis. In a meta-analysis of amniotic products, 4/5 trial protocols were designed to change the product weekly. In the fifth trial where changes were left to provider discretion, there was no decease in wound healing.³⁰

Societal Input

NICE Diabetic foot problems: prevention and management¹⁶⁴

The clinical guideline on diabetic foot problems considers dermal or skin substitute grafts as an appropriate addition to standard care in treating DFU only when healing has not progressed with SOC on the advice of the multidisciplinary foot care service.

International Working Group on the Diabetic Foot (IWGDF)¹²

IWGDF recommends the consideration of placental-derived products as an adjunctive treatment to the best SOC when standard care alone has failed to reduce the size of the ulcer. (GRADE Strength of recommendation: Weak; Quality of evidence: Low). This was based on several studies, including those of moderate bias, suggesting that placenta-derived products may have a beneficial effect on ulcer healing. The authors also state these findings need to be confirmed in large, randomized trials and there is insufficient evidence to support superiority of any product(s).

For topically applied treatments, the IWGDF advises against the use of bioengineered skin products compared to SOC.

For both recommendations, the IWGDF considered the available evidence to be of low quality, and their recommendation was weak (e.g., based on the quality of evidence, balance between benefits and harms, patient values and preferences, and cost or resource utilization).

Wound Healing Society (WHS)^5,11

The WHS has published updated evidence-based guidelines on the treatment of diabetic ulcers. Regarding the use of skin substitutes, the WHS concluded that level I evidence suggests that cellular and acellular skin equivalents improve the healing of diabetes-related foot ulcers. In these guidelines Level I required at least 2 RCTs supporting the intervention of the guidelines. The quality of evidence was not assessed.

• In evidence-based guidelines for venous ulcers, the WHS stated that there is evidence that a bi-layered living HSE, used in conjunction with compression bandaging, increases the incidence and speed of healing for venous ulcers compared with compression and a simple dressing (Level I evidence). The WHS recommends adequate ulcer bed preparation and control of excess bioburden levels prior to application of a biologically active dressing.

They also noted that cultured epithelial autografts or allografts have not been demonstrated to improve stable healing of venous ulcers (Level I). The WHS also stated that there is Level II evidence that a porcine small intestinal submucosal construct may enhance healing of venous ulcers.⁵

Society for Vascular Surgery/American Podiatric Medical Association/Society for Vascular Medicine (SVS/APMA/SVM)

The SVS/APMA/SVM published a joint evidence-based guideline using GRADE system for the management of patients with diabetes, including treatment of diabetes related chronic foot ulcers.⁷

These organizations’ recommendations for DFU failing to demonstrate improvement (> 50% ulcer area reduction) after a minimum of 4 weeks of standard ulcer therapy include:

• Adjunctive ulcer therapy options with negative pressure therapy, biologics (platelet-derived growth factor, living cellular therapy, ECM products, amniotic membrane products) and hyperbaric oxygen therapy. The choice of adjuvant therapy is based on clinical findings, availability of therapy, and cost-effectiveness; there is no recommendation on ordering of therapy choice (Grade 1B).
• Consideration of living cellular therapy using a bilayer keratinocyte/fibroblast construct or a fibroblast-seeded matrix for treatment of DFU when the individual is recalcitrant to standard therapy (Grade 2B).
• Consideration of the use of ECM products employing acellular human dermis or porcine small intestinal submucosal tissue as an adjunctive therapy for DFU when the individual is recalcitrant to standard therapy (Grade 2C).

Wound Healing Foundation (WHF)

The WHF published the results of a Consensus Panel on Chronic Wounds composed of dermatology, general surgery, vascular surgery, pediatric surgery, plastic surgery, podiatry, nursing, and wound healing research experts in diverse practice settings. The panel agreed that a chronic wound is designated as a “stalled wound” when it has failed to progress towards healing, following 4 weeks of standard evaluation and management during which identified etiologic factors have been addressed. The importance of treating the underlying condition contributing to the wound development is emphasized as essential for healing. Identified elements in the SOC treatment for these wounds include debridement, infection control, moisture management, dressing and protection, compression in venous and lymphatic ulcers, and offloading. Negative pressure wound therapy, grafting and hyperbaric oxygen are identified as advanced or adjunctive treatment modalities. Decision-making depending on the level of evidence for a specific product and wound type is recommended for CTP. Unlike autologous skin grafts, the hom*ologous grafts do not persist and act as a template for cell growth; however, advantages include no donor site, application in office or operating room, possible growth factors and immunomodulators, reduction of insensible water loss and preparation of wound bed for autografting. Disadvantages include prolonged or repeat applications which may delay final grafting and definitive wound coverage. However, the consensus panel did not include the evidence level or qualify the strength of this recommendation.¹⁰

Analysis of Evidence (Rationale for Determination)

Due to the heterogeneity of RCTs, poor study designs, small sample sizes, lack of comparators or standardization of practices, lack of long-term efficacy and safety data, and high risk of bias in the current body of literature, there is insufficient evidence to demonstrate efficacy of most skin substitute/CTP in DFU/VLU healing. Moreover, this evidence is challenged by a low level of certainty in the estimated wound-healing effect attributable to these products. Many products have been marketed as substantial equivalents without establishing their role in ulcer healing. Potential risks with these products are not adequately addressed due to lack of long-term safety. Lastly, clinical outcomes have rarely been reported beyond 12 weeks in the current literature, raising additional concern for the durability of the estimated therapeutic benefit(s).

Despite these limitations, a promising trend within the literature towards outcome improvement is identified. Therefore, to be considered reasonable and necessary for coverage, each skin substitute/CTP must demonstrate net positive health outcome(s) in a well-defined patient population. Specifically, wound closure attributable to the individual product(s) proven in clinical trials with meaningful degree of certainty is required. Therefore, a limited coverage position for specific products in specific patient populations has been taken to facilitate access to skin substitute graft/CTP products with clinically meaningful net-positive clinical outcome(s) validated by evidentiary review.

The intent of a skin substitute graft/CTP is to augment wound healing by promotion of skin growth and wound closure. Inherent to this process is stability and adherence of the product which allows it to remain in place to promote skin growth and wound closure with incorporation of the graft. A product requiring removal or replacement without the benefit of incorporation more clearly is characterized as a dressing. There is a trend within the published literature suggesting that products with fewer applications result in shorter closure time. However, direct comparisons of products have not been conducted. Most products resorb into the wound, therefore additional product may be beneficial to facilitate continued wound closure in the event the wound is improving with the use of the skin substitute graft. While there are no studies that directly assess the number of applications required for wound closure, reports throughout the literature suggest the mean closure time is approximately 4 applications within 12 weeks. In the largest reported cohort of 12,313 ulcers treated with skin substitute grafts the mean number of applications was 3.7 with a standard deviation of 3.6.¹⁷ Moreover, products evaluated in the evidence review also reported a similar number of applications and time duration. Based on this evidence most ulcers would be expected to close within a maximum of 4 applications and 12 weeks. Ulcer size and immune compromise has been cited by expert opinion as grounds for additional applications secondary to extended time to heal. Therefore, an exception has been added to the LCD to ensure that patients who have documented benefit of wound healing with the skin substitute graft use may receive additional applications or duration of care with documented clinical indications. The additional application or extension would be identified with a modifier and documentation in the medical record will be required to explain the clinical rationale for the exception and may be subject to medical review.

There is a clear need for further investigation and understanding of skin substitute grafts and their role in management of chronic wounds such as DFU and VLU. Future investigations will clarify the role of these products, compare products, establish standardized practice for utilization and allow a better understanding of products (and alternative treatments) most beneficial to healing diverse wounds, with the expectation of improved outcomes for patients suffering from these complex conditions.