Amniotic Membrane Adjuncts and Clinical Applications in Wound Healing: A Review of the Literature

Introduction. Recent advances in the preservation and processing of amnion/chorion tissue have dramatically increased the bioavailability of these wound healing factors as well as the shelf life of their related tissue products, allowing for a surge in clinical use. Many studies, including basic science, clinical trials, and randomized controlled trials, have emerged examining the biologic properties of amnion/chorion membrane products and their efficacy in wound healing. Objective. A literature review was conducted regarding the safety and efficacy of amniotic membrane adjuncts. Methods. The PubMed and MEDLINE databases were queried and sorted based on clinical trials with publication dates ranging from 2013 to 2017. Only studies pertaining to human subjects were included for review. Results. Amnion/chorion membranes have been studied in the treatment of burns, diabetic foot ulcers, fistulas, ocular defects, and venous leg ulcers, among other wounds. Amnion/chorion allografts were found to be beneficial in the setting of difficult-to-heal fistulas and were effective in treating diabetic and venous ulcers when combined with standard therapy. Conclusions. Overall, clinical trials have demonstrated that patients treated with amniotic membrane products have increased rates of wound healing compared with the standard of care. Additional trials are needed to examine more amnion/chorion membrane products.

Wound healing is an intricate and complex process, with numerous physiologic mechanisms occurring in a coordinated and synchronized manner. With respect to failure of wound healing, any individual component or combination of the physiologic processes may be impaired. Attempting to target a single entity is unlikely to substantially and clinically progress wound healing. As such, the ideal therapeutic product would target multiple processes and enhance deficiencies in the system. Human amnion/chorion tissue products have been researched and developed to address the lack of a multifaceted tissue regeneration and wound closure product.^1-7 Human amnion/chorion tissue products are derived from the placenta and contain numerous growth factors and cytokines that have proven to promote wound healing.^6,7 Despite advances in the understanding of the role of amnion/chorion tissue in wound healing, preserving the biologic activity of these allografts has been a major hurdle in their widespread clinical use. However, significant progress has been made in this arena, and 5 manufacturers have developed processes to retain the biologic activity of these amniotic tissues.^2,4,6 Over the preceding decade, a wide range of basic science and clinical studies have been published^1-7 examining both the in vitro and in vivo properties of human amnion/chorion products. This article presents a comprehensive overview of the literature and provides background information on the fundamental biology of human amnion/chorion products as well as current evidence supporting their use and efficacy in clinical practice.

The amniotic membrane is found in the inner layer of the placenta and is composed of both amnion and chorion membranes.⁷ The amniotic membrane is a thin, highly durable biomaterial;  it supports the developing fetus and amniotic fluid for the entirety of the pregnancy.⁸ Amniotic membranes are avascular and lack a direct blood supply,⁹ and all nutrients are supplied via diffusion from the amniotic fluid or the underlying decidua.¹⁰ The amniotic membrane also secretes substances into the amniotic fluid and uterus, contributing to the homeostasis of amniotic fluid and maternal cellular physiology.¹¹

Amnion is composed of an epithelial layer, a basement membrane layer, a compact layer, a fibroblast layer, and an intermediate layer.⁷ Each of these 5 layers contain a unique arrangement of a variety of cells, collagen types, and structual proteins; together, they provide the mechanical integrity of the membrane.^7,10 In comparison, chorion is approximately 4 times thicker than amnion and composed of a reticular layer, a basement membrane layer, and a trophoblast layer.^6,8,9

The amniotic membrane contains numerous growth factors, cytokines, and signaling molecules that play important roles in fetal development and gestation.⁷ Interestingly, these molecules also are critically important in tissue regeneration and wound healing. Enzyme-linked immunosorbent assays of EpiFix, a dehydrated human amnion/chorion membrane (dHACM; MiMedx, Inc, Marietta, GA) have demonstrated quantifiable levels of vascular endothelial growth factor, platelet-
derived growth factors AA and BB, transforming growth factors alpha and beta, basic fibroblast growth factor, epidermal growth factor, and granulocyte-colony stimulating factor.^6,7 In addition, interleukins 4, 6, 8, and 10 were detected in the samples, and it is hypothesized that their immunosuppressive functions may play a role in the immune-privileged properties of amniotic membrane products such as dHACM.⁶

Amnion is nonimmunogenic and has been shown to reduce inflammation and pain while also serving as a matrix for cell deposition.⁶ In recent years, the use of human amniotic membrane products has increased exponentially in the clinical setting. Ophthalmic surgeons have used amnion/chorion membrane products to treat corneal ulcerations, conjunctival lesions, and chemical burns of the eye.¹² Recent studies^1-4 have examined the utility of amnion/chorion membrane products in chronic cutaneous wounds, including diabetic foot and venous leg ulcers (Tables 1, 2).

The PubMed and MEDLINE databases were queried for the terms human amnion chorion membranes, dehydrated amnion chorion, amnion chorion membrane wound healing, diabetic foot ulcer, and venous ulcer. Articles were sorted based on type, and clinical trials were selected for review for inclusion in analysis. The publication dates of included articles ranged from 2013 to 2017. Articles were excluded if they only reported basic science or animal model data. The data collected included wound type, study design, study size, wound care products used, and measured study outcomes.

Clinical applications and studies of amnion/chorion for wound healing

Chronic lower extremity diabetic ulcers. In 2013, Zelen et al² published one of the earliest randomized trials comparing dHACM products against the standard of care (SOC) in the treatment of nonhealing diabetic foot ulcers (DFUs). The study was a prospective, stratified, randomized, comparative, parallel-group, nonblinded clinical trial of dHACM versus standard protocol moist wound care therapy. The study was conducted at a single center.

Eligible patients (N = 25) had DFUs between 1 cm² and 25 cm² in size that were of > 4-weeks’ duration. Patients randomized to the SOC group (n = 12) underwent wound debridement and moist wound therapy with SilvaSorb Gel (Medline Industries, Inc, Mundelein, IL) and AQUACEL Ag (ConvaTec, Bridgewater, NJ). Patients in the dHACM group (n = 13) received an application of the allograft after surgical debridement of necrotic tissue. The dHACM was covered with a nonadherent dressing (ADAPTIC TOUCH Non-Adhering Silicone Dressing; Acelity, San Antonio, TX), a moisture-retentive dressing (hydrogel bolster), and a compression dressing. The SOC group underwent daily dressing changes, whereas the dHACM group underwent weekly dressing changes.²

Patients were seen by the investigator prior to the start of therapy and at least once every 7 days for up to 12 weeks or until complete healing. Patients were removed from the study if they did not achieve at least 50% wound area reduction at 6 weeks. The primary study outcomes included reduction of wound size and percentage of complete wound healing at 4 and 6 weeks. A final wound evaluation was performed at 12 weeks for patients remaining in the study.²

At week 4, the average ulcer surface area reduction was 97.1% ± 7.0% in the dHACM group and 32.0% ± 47.3% in the SOC group (P < .001). At week 6, 10 of 12 (83.3%) patients in the SOC group chose to leave the study, as they failed to see improvements in wound healing. Therefore, analysis was carried out from data collected at the 6-week mark. At 6 weeks, 92.3% (12/13) of patients in the dHACM group had complete wound healing, whereas only 8.3% (1/12) of patients in the SOC group had complete wound healing.

The dHACM products also were evaluated for their efficacy in healing chronic lower extremity ulcers in patients with diabetes in a prospective, randomized, controlled, multicenter study, with SOC serving as the comparison.³ Sixty patients were enrolled in the study and randomized (20 patients/group) into receiving a weekly application of Apligraf (bioengineered skin substitute, BSS; Organogenesis, Inc, Canton, MA), dHACM, or SOC with collagen-alginate dressing. The primary objective was to determine the percentage of patients in each treatment arm who had complete wound healing at 4 and 6 weeks of treatment. At week 4, 35% (7/20) of patients receiving BSS, 85% (17/20) of patients receiving dHACM, and 30% (6/20) of patients receiving SOC had complete healing. At week 6, 45% (9/20) of patients receiving BSS, 95% (19/20) of patients receiving dHACM, and 35% (7/20) of patients receiving SOC achieved complete wound healing. At both 4 and 6 weeks after treatment initiation, dHACM had significantly increased rates of wound healing compared with BSS and SOC.

Zelen et al¹³ continued their comparison of dHACM with BSS and SOC with an additional study of 40 patients. Using this 100-patient cohort, the authors compared clinical outcomes at 12 weeks. Patients were randomized to receive BSS, dHACM, or SOC with collagen-alginate dressing. At week 12, the proportion of wounds achieving complete closure was 73% (24/33) for patients receiving BSS, 97% (31/32) for patients receiving dHACM, and 51% (18/35) for patients receiving SOC (adjusted P = .00019). Patients treated with dHACM had a higher probability of wound healing (hazard ratio [HR], 5.66; adjusted P = 1.3 x 10^-7) compared with SOC alone.¹³ No statistically significant difference in the probability of wound healing was measured between the BSS and SOC groups. However, patients treated with BSS were less likely to achieve wound healing than patients treated with dHACM (HR, 0.30; 95% confidence interval, 0.17–0.54; unadjusted P = 5.8 x 10^-5). In addition, patients required an average of 6 BSS grafts to achieve healing, whereas only an average of 2.5 dHACM grafts were needed. This corresponded with a median graft cost of $8918 (range, $1458-$19 323) for BSS and $1517 (range, $434-$25 710) for dHACM.

DiDomenico et al¹⁴ compared aseptically processed dehydrated human amnion/chorion allografts (dHACA; AmnioBand; Musculoskeletal Transplant Foundation, Edison, NJ) versus SOC in wound closure of nonhealing DFUs. The primary objective was to compare wound healing rates after 6 weeks of dHACA application as an adjunct to SOC versus SOC alone. Patients with unhealed DFUs who failed conservative therapy for at least 4 weeks (N = 40) were randomized 1:1 to receive either dHACA plus SOC or SOC alone. Wounds included in the study were between 1 cm² and 25 cm² in size. The study was conducted at 5 wound care centers in the United States. All eligible wounds were managed for a 2-week period with SOC alone prior to randomization. Patients were included in the study only if the DFU did not reduce in size by more than 20% at the end of the screening period.

Patients were treated weekly until wound closure or for 12 weeks. Treatment with SOC entailed daily dressing changes with collagen alginate (FIBRACOL Plus Collagen Wound Dressing with Alginate; Acelity). Wounds were dressed at home 6 days per week and by the site investigator 1 day per week. In the dHACA plus SOC group, wounds were traced on the graft and the graft was cut to size. The graft was covered with a nonadherent dressing (same as Zelen et al²), a moisture-retentive dressing, and a padded 3-layer dressing (Dyna-Flex; Systagenix, an Acelity company, San Antonio, TX). Patients in the dHACA plus SOC group received dressing changes once weekly. Six weeks after randomization, all wounds were examined and the percentage area reduction was calculated. Patients were withdrawn from the study if the percentage area reduction was < 50%.

The primary endpoint was a comparison of wound healing at 6 weeks between the 2 treatment arms. Some of the secondary endpoints included proportion of wounds healed at 12 weeks, time to heal, number of graft applications, graft waste, and cost of graft for healed wounds. The 2 groups were well matched; however, the index wound in the SOC group was larger than in the dHACA plus SOC group (3.3 cm² vs. 2.01 cm², respectively). One patient in the SOC group was lost to follow-up due to a serious adverse event. No patients were excluded from analysis in either group.

At week 6, 70% (14/20) of the wounds in the dHACA plus SOC group had healed compared with 15% (3/20) in the SOC group (P = .001), and mean time to heal was 30 days in the dHACA plus SOC group versus 40 days for the SOC group. At week 12, mean time to heal was 36 days in the dHACA plus SOC group compared with 70 days in the SOC group. At week 6, the mean number of grafts used per wound was 3.1 (± 1.7), and the mean cost of the graft to heal ulcers was $1091 (± $619; n = 14). Of note, 8 patients were withdrawn from the SOC group and 1 from the dHACA plus SOC group because their wounds failed to reduce in area by at least 50%. No patients in the dHACA plus SOC group had any adverse events related to the graft. A summary of these studies is provided in Table 1.

Venous leg ulcers. Venous leg ulcers (VLUs) are the most common type of lower extremity wound, and upward of 80% of leg ulcers involve the venous system.¹⁵ It is estimated that 500 000 to 2 million Americans develop chronic VLUs, with the prevalence expected to further rise with the aging population as well as the increased rates of obesity and congestive heart failure.^16,17 Chronic VLUs generally require weeks to months to heal, and patients face significant pain and impairment of function during this process. Graduated compression bandaging is the SOC for chronic VLUs; however, a significant need exists for a product allowing for a more rapid and timely healing. The utility of dHACMs in the treatment of VLUs was examined in 2014 in a randomized, controlled clinical trial in which dHACM plus SOC were compared with SOC alone.⁴ This was the first randomized trial analyzing the safety and efficacy of dHACM as a treatment for VLUs. The primary outcome was the percentage of patients who achieved 40% wound closure at 4 weeks. Eighty-four patients were enrolled in the study; 63% (53/84) were randomized to the dHACM plus SOC group and 37% (31/84) received SOC alone. At week 4, 62% of patients in the dHACA plus SOC group and 32% in the SOC alone group showed > 40% wound closure (P = .005). Also, the dHACA plus SOC group had a mean size reduction of 48.1% compared with 19.0% in the SOC alone group.⁴

In 2015, Serena et al¹⁸ published a follow-up study in which they retrospectively examined whether the results of their original 4-week study correctly correlated with rates of complete wound healing at 24 weeks. The records of 44 patients who had failed to achieve complete wound healing during the initial trial were included. Twenty patients (45.4%) had wounds that reduced in size > 40%, and 24 patients (55%) had a reduction of < 40% during the initial study. Of the group with > 40% healing at week 4, 80% (16/20) achieved complete healing in a mean of 46 days, whereas 33.3% (8/24) of the < 40% healing group achieved complete healing in a mean of 103.6 days. This study demonstrated a true correlation of healing between the 4-week and 24-week trials in 73% (32/44) of patients.

As of September 2017, only 2 clinical trials^4,18 have been published examining the efficacy of dHACM in the setting of lower extremity venous ulcers (Table 2).

Other methods of amnion/chorion preservation

Although the majority of the relevant literature has reported on the use of dHACM products, cryopreserved human amniotic membrane products (eg, NEOX; Amniox Medical, Atlanta, GA) are available. The previously mentioned manufacturer has posted 8 case reports regarding their cryopreserved human amniotic membrane product on their website¹⁹; however, no peer-reviewed publications have been published on this product and thus were not included in this manuscript.

Perepelkin et al²⁰ compared the viability and posttransplant outcomes of cryopreserved amniotic membrane with hypothermically stored amniotic membranes. The authors harvested freshly delivered placenta, manually separated the amniotic membrane, and stored the membranes in a cryopreservation solution containing X-VIVO 10 media (Lonza, Walkersville, MD) and 10% dimethyl sulfoxide. Fluorescent nuclear dyes were used to differentiate between viable and dead cells. The authors reported no statistically significant difference in the percentage of viable cells in whole tissue segments between the hypothermically stored membranes compared with the cryopreserved amniotic membranes. A slight statistically significant increase in viable cells was observed in the epithelial layer of the hypothermically stored membrane compared with the cryopreserved amniotic membrane.²⁰ The authors provided cryopreserved amniotic membranes to surgeons for use in a variety of clinical applications and reported that 16% (8/49) of these membranes were rated excellent and 82% (40/49) were “satisfactory/fair/good, which caused minimal scarring and promoted a good healing response.” The remaining 2% (1/49) was deemed unsatisfactory, but the indicated reason was neither related to the allograft nor classified as an allograft-associated adverse event.²⁰ Without the use of a standard measure of successful wound healing and proper controls, such data provide little insight into the clinical benefit of cryopreserved amniotic membrane products over dHACM products or SOC.

The application of human amniotic membranes in wound healing dates back to at least the 1940s, when ophthalmologists began exploring their utility in persistent corneal defects.²¹ The numerous cytokines and growth factors found in the placenta make amnion/chorion membranes advantageous for wound healing and tissue regeneration.^6,7,10 Although it is possible to harvest amnion/chorion membranes from patients undergoing elective cesarean section, the development of technical advances in the processing and preservation of dHACM allows for the grafts to be stored for several months, enabling their application in both emergent and elective procedures.^2,4,6

Both fresh and preserved amnion/chorion membranes have been studied in the treatment of a wide variety of wounds including burns, diabetic ulcers, fistulas, ocular defects, and venous ulcers, with studies ranging from single case reports to randomized controlled trials.^3,4,22-26 Amnion/chorion allografts provide substantial benefit in the setting of difficult-to-heal fistulas, with no recurrence of the defect.^22,25,26 With respect to both diabetic ulcers and venous ulcers, the use of amnion/chorion membranes in addition to SOC was more effective than standard therapy alone.^2-4,13,14 As a whole, patients in the dHACM treatment arm had increased rates and shorter durations to complete wound healing.

Although the data from these clinical trials are promising, it must be noted that the trials are industry supported. The allografts remain costly; therefore, researchers are unable to conduct such trials independently. In addition, most of the studies have focused on dHACM, and little has been published on competing amniotic membrane products. Generating supportive evidence for the clinical utility of dHACM is necessary; however, additional trials examining other dHACM products would be beneficial, both to increase the general knowledge of dHACMs and to allow clinicians to make evidence-based decisions in the selection of wound care products.

Another potential shortcoming of these studies is the brevity of the trial period. The studies reported here examined outcomes between 4 and 6 weeks. Wound healing is often incomplete at this time point. However, clinical trials are a costly endeavor, particularly when studying new biologic wound healing products. Ideally, these studies would continue for 14 to 20 weeks, but this is an impractical standard and one that is unlikely to be met. The 4-week to 6-week period has been widely used in the study of wound care products.

Clinicians also may question the real-world correlation of cytokines and growth factor levels in the dHACMs measured in vitro and after they are placed in a chronic wound bed. The inflammatory environment of a chronic wound is full of metalloproteinases that can degrade the proteins in dHACM. The importance of developing a clean wound bed, with careful debridement of necrotic tissue, bacteria, and enzymes, cannot be overstated. This principle is true in wound care, but it is especially important when attempting to treat chronic wounds with biologic materials. To provide compelling evidence of the bioactive mechanism of dHACMs, future trials should consider quantifying levels of cytokines and growth factors in the wound bed throughout the course of dHACM treatment.

Wound healing is an intricately complex process requiring the coordination of numerous physiologic processes. In both chronically ill and acutely injured patients, these processes may be compromised, resulting in impaired wound healing and tissue regeneration. The theoretical benefit of amnion/chorion membranes has long been hypothesized, but recent studies have begun to show a practical benefit to patients. A handful of clinical trials have been performed that examine the safety and efficacy of dHACM in the setting of diabetic and venous lower extremity ulcers. These studies demonstrate an increased rate of wound healing for patients treated with dHACM over SOC and bilayered skin substitutes.

Affiliations: Rutgers Robert Wood Johnson Medical School, New Brunswick, NJ; The Ohio State University Medical Center, Department of Plastic and Reconstructive Surgery, Columbus, OH; and Rutgers New Jersey Medical School, Newark, NJ

Correspondence: Mark S. Granick, MD, Rutgers New Jersey Medical School, 90 Bergen Street, Suite 7200, Newark, NJ 07109; mgranickmd@njms.rutgers.edu

Disclosure: The authors disclose no financial or other conflicts of interest.