Real-world Experience With a Decellularized Dehydrated Human Amniotic Membrane Allograft

While randomized controlled trials (RCTs) are designed to evaluate efficacy and/or safety under controlled conditions, use of strict inclusion/exclusion criteria are noted to exclude more than 50% of wound populations. Applicability of RCT outcomes to performance expectations in real-world wound populations raises questions about generalizing their results. The primary aim of this decellularized, dehydrated human amniotic membrane (DDHAM) Use Registry Study was to gain experience and observe outcomes with use of a DDHAM in uninfected, full-thickness, or partial-thickness wounds that, in the investigators’ opinions, would benefit from such treatment.   Methods. Investigators were instructed to provide usual care regarding visit and application frequencies, concomitant therapies, and change in wound-care regimens. The only exclusions were patients with actively infected wounds or known hypersensitivity to DDHAM. Fifteen sites with practicing wound care clinicians of various specialties participated in this review, enrolling chronic wounds including venous, diabetic, pressure, collagen vascular, and arterial ulcers—all of various severities, durations, sizes, and previous treatments. Twenty-eight ulcers studied had failed 32 previous treatments with advanced biologic therapies. A total of 244 wounds were observed in this study, however, this review is limited to the 179 chronic wounds in 165 patients that were enrolled at 15 of the 19 participating centers. The 4 centers that enrolled acute wounds only were excluded. Results. Results from the analysis of this very heterogeneous population demonstrated that during the usual course of an average of 8 weeks of wound management, patients experienced factors that significantly affected wound closure. These factors included wound infections, noncompliance with prescribed treatments (eg, compression, off-loading, and wound care), re-injury of the wound, and systemic comorbidities. Nearly 50% of chronic wounds (including those that failed previous therapy with advanced biologics) with an average baseline area of 3.1 cm² achieved complete closure within a median of 6.3 weeks without product-related adverse experiences. Conclusion. Despite the challenges of uncontrolled factors that affect healing, this registry study demonstrated the safety and clinical benefit of DDHAM to support wound closure across a variety of chronic wound types and patient conditions in real-world environments.

Decellularized, dehydrated human amniotic membrane (DDHAM) allograft (Biovance, Alliqua Biomedical, Langhorne, PA) is regulated under Section 361 of the Public Health Service Act indicated for the management of uninfected partial-thickness and full-thickness acute and chronic wounds.¹ This allograft represents amniotic membrane that underwent minimum processing designed to clean, store, and preserve the tissues without altering the native architecture. The membrane is composed of a complex extracellular matrix that is dehydrated and terminally sterilized with electron beam irradiation. It can be stored at room temperature and has a shelf life of 5 years.

Since 1910, the use of human amniotic membrane in burn and wound management has been described in medical literature,^2-4 although a larger interest in the clinical applications of amnion did not develop until the 1970’s, when fresh, rinsed amniotic membrane was used on full-thickness wounds. Biopsies of second degree burn wounds treated with dehydrated decellularized amniotic membrane allograft taken at 5 days showed complete healing—even a thin stratum corneum had developed (Treadwell, unpublished data, 2005). As added benefits, reduced pain,^5-7 fluid loss,^8,9 and a reduction in infection was reported in association with its use.^10,11 In the 1980’s, fresh amnion was reported as beneficial for the treatment of chronic wounds.^12,13

With the advent of the need to prevent passage of infectious diseases via allografts,^14,15 use of fresh amnion became less common. This sterile DDHAM allograft was produced to provide a replacement for fresh amnion that minimized the potential for infection. In addition, cells and cellular debris are removed to reduce the potential for sensitization or immune reactions that may result with exposure to any residual maternal cellular material. This is accomplished while maintaining the benefits of amnion for the support of wound healing. A registry study was conducted to obtain proof of this principle and explore the wound types that would most benefit.

Prospective, randomized, clinical trials are designed to test the effect of a specific intervention. To limit heterogeneity of the study population and minimize the size of the population needed to interpret the effect, strict inclusion and exclusion criteria are placed in an attempt to decrease variability that may dilute the test treatment effect.¹ In the process of applying limits through inclusion and exclusion criteria to create as homogeneous a test population as possible, more than 50% of the wound care population typically seen is excluded.¹⁶ A registry that includes all types of wounds without excluding a vast majority of enrolled patients provides a more realistic test of effect of an intervention for patients, providers, and payors. However, it is clear that the traditional outcome of percent of wounds with complete closure may be greatly diluted, as the many factors that are usually excluded may have an impact on that outcome.

A registry study was conducted to obtain evidence of the clinical benefit of using a sterile, DDHAM allograft and to explore the wound types that would most benefit. The data reported from these study observations were analyzed to determine safety and performance when DDHAM was applied on a variety of acute and chronic wounds. The registry was conducted in 2005. The investigators chosen for participation were located across the continental United States and were experienced wound care practitioners. The basic premise of good wound care practices included debridement, maintenance of a moist environment, and pressure-offloading or compression where needed based on wound type, in line with the now commonly accepted TIME approach (tissue, infection/inflammation, moisture balance, and edge of wound) to wound assessment and wound bed preparation.^17,18 Advanced wound therapies available included topical recombinant human platelet-derived growth factor, living skin substitutes, acellular human dermal matrix, and a porcine small intestinal submucosa wound matrix. Decellularized, dehydrated human amniotic membrane was positioned to be a new class of advanced wound therapy.

The study included both chronic and acute wounds. While the acute wound population experienced similar or greater clinical benefits from the use of DDHAM in their treatment regimens, this report will concentrate on the largest population—patients with chronic wounds.

Objective. The DDHAM Use Registry Study was designed to gain extensive clinical and safety experience in the use and performance of DDHAM in the treatment of any uninfected, full-thickness, or partial-thickness wound. This report is limited to the chronic wound population.

Design. Any subject with a chronic wound who, in the investigator’s opinion, would benefit from treatment with DDHAM was eligible for the study with the exceptions of those with a known sensitivity to DDHAM and/or wound infection present at time of screening. Actively infected ulcers (diagnosed on the basis of clinical signs and symptoms) were permitted to be enrolled after appropriate treatment. The concern was that proteases released during active infections¹⁹ would likely digest applied allografts, similar to proteases effect on host tissues. The criterion was placed to avoid wasting allograft. This exclusion criterion is congruent with other commercially available allografts where infected wounds are a contraindication for allograft application.²⁰

This study was observational; there was neither treatment randomization nor blinding, and there were no control groups. Subjects had to provide written consent to participate and the informed consent form and protocol received appropriate study approvals prior to initiating enrollment at each study site.

Investigators were chosen based on their clinical experience with wounds. There were no prohibited prior medications and only 1 prohibited concomitant therapy—topical enzymatic debriding agents (ie, collagenases). Frequency of reapplication of DDHAM was at the discretion of the investigator based on their subjects’ responses to the therapy. There were no predetermined reasons for removing subjects from therapy with DDHAM other than complete wound closure. Investigators were given the liberty to decide when participation of a specific subject should discontinue. Therefore, the length of time for individual subject participation was determined by the investigator or the subject.

Data collected during the study included wound length (L, the longest dimension of the wound) and width (W, perpendicular to length at the widest point), surface area (calculated using the formula for an ellipse [L/2 x W/2 x pi), related comorbidities and medical history, wound care regimen notes, recorded subject observations, number of applications of DDHAM, and photographs taken at baseline and final treatment visit.

Data were collected on paper case report forms from the first and last subject visits and submitted for entry into a spreadsheet database (Excel Microsoft Corporation, Redmond, WA).  Many of the responses were contained in comment fields verbatim. During the analysis, only those factors that were listed as present were counted. If a comorbidity was present but not recorded in the data collection form, it was considered as absent in the subject for the study. In some cases, generalizations like “general poor health” or “serious comorbidities” were submitted in the record, but these could not be assigned to specific comorbidities or factors that may have affected wound closure. However, all subjects whose wounds were categorized as diabetic foot ulcers were assumed to have diabetes mellitus as a comorbidity.

Investigators were instructed to provide their usual standard of wound care to the subjects regarding frequencies of visits, allograft applications, and concomitant therapies. The decellularized, dehydrated human amniotic membrane had to be used in line with the package insert which states that the allograft be placed on debrided wounds that are not infected and that collagenases should not be used in combination with the allograft. Even though subjects may have remained under the investigators care, once DDHAM was discontinued, further information about the subject’s clinical course was not provided.

All tabulations were based on the observed or recorded data; no results were imputed. Continuous variables were summarized using descriptive statistics and compared between closed vs unclosed wounds by wound type and overall, independent of wound type, using a 1-factor analysis of variance test. Categorical variables were summarized using counts, percentages, and exact 95% binomial confidence intervals and compared between closed vs unclosed wounds by wound type and overall, independent of wound type, using a 2-tailed Fisher’s exact test or 2-sample Wilcoxon test.

Fifteen of the 19 centers that participated collected data on the treatment of chronic wounds (4 of the 19 enrolled only acute wounds and were excluded from this study) between December 2004 and September 2005. Based on the data analyses, 2 sub-cohorts of the chronic wounds were defined. The Intent-to-Treat (ITT) group included the set of wounds of subjects who had initial and final measurements and a treatment/observation period greater than 3 days. The Good Wound Care (GWC) Group comprised a large subgroup within the ITT group that resulted from removing the wounds of subjects who were noted to be grossly noncompliant with the prescribed wound care regimen. This type of noncompliance included failure to use offloading in cases of diabetic foot ulcers or pressure ulcers; failure to use compression therapy if a venous stasis ulcer was present, or removal of the allograft by the subject after it was placed. Also excluded were protocol violations, particularly the presence of an active wound infection or underlying osteomyelitis at the time of initial allograft application, as these subjects should not have been enrolled. The GWC group totaled 121 chronic wounds.

A total of 165 subjects with 179 chronic wounds were enrolled in the study. Of the total number of subjects, 98 (59.5%) were male and 67 (40.5%) were female (Table 1). The most common types of wounds were venous leg ulcers (n = 98, 49.7%), followed by diabetic foot ulcers (n = 47, 26.3%), pressure ulcers (n = 20, 11.2%), arterial ulcers (n = 15, 8.4%), and ulcers caused by a collagen vascular disease such as lupus or rheumatoid arthritis (n = 8, 4.5%). (Percentages may not add up to 100 due to rounding.)

In the ITT group (n = 154), 11 (6.7%) subjects were withdrawn for either not having baseline or final wound measurements because of death or unknown reasons; withdrawing consent; having less than 3 days of observation prior to healing; or being lost to follow-up. Table 2 provides the consort statement by wound type.

The average age of all subjects at the first visit was 64.4 years with 71 of the 154 subjects (46.4%) being older than 65 years. In the ITT group, 22 (14.3%) subjects were 80 years of age or older, and 10 (6.5%) were older than 85 years. Age was not a significant determinant of wound closure whether it was analyzed as a continuous variable (P = 0.8009) or when subjects greater than 75 years old were compared to those less than or equal to 75 years of age (P = 0.8554) for closed vs unclosed chronic wounds.

Wound size distribution was divided into ranges, < 2cm²; 2 – ≤ 10cm²; > 10 – ≤ 15cm²; >15 – ≤ 25cm²; and > 25cm² (Table 2). When all ulcer etiologies were analyzed together, most lesions were in the 2 - ≤ 10 cm² range (n = 75, 42.1%). This was also the case for arterial ulcers (n = 6, 40.0%), diabetic foot ulcers (n = 20, 42.6%), and venous leg ulcers (n = 37, 42.0%). For pressure ulcers, most ulcers were < 2 cm² (n = 10, 50.0%). Ten of the 16 (62.5%) ulcers that were > 25 cm² in area were of a venous stasis etiology (20.3% of the 79 venous ulcers; Table 3). The baseline wound area was a significant determinant for closure of chronic wounds (P = 0.0043).

The majority of chronic ulcers were located on the lower extremities (162/166, 98%). Therefore, there was no difference in the clinical effectiveness of DDHAM when analyzing outcomes in chronic lower extremity wounds compared to the overall ITT group in this heterogeneous patient population. However, controlling for gross noncompliance, the GWC wound population demonstrated an approximately 20% increase in the wound closure incidence compared to the ITT population (Table 4). Because there were only 4 ulcers located outside of the lower extremity, analysis on the basis of location of the ulcer was not undertaken.

Table 5 provides a list of comorbidities and factors that could potentially affect healing.¹⁹ On average, records noted a mean of 2.02 of these factors/comorbidities, with diabetes mellitus being the most common one (n = 65, 39.4% of subjects), followed by arterial insufficiency (n = 34, 10.2%). When all of the potential factors that may have a negative effect on healing are considered together, it is apparent that subjects with more than 1 factor were less likely to achieve complete wound closure, and that the probability of closure is inversely related to the number of factors—whether or not good wound care was evidenced (Figure 1). The number of factors was a significant determinant of closure (P = 0.0027). Subjects with no risk factors that may have a negative impact on healing were 4.3 times more likely to achieve wound closure compared to patients with 3 or more factors (P = 0.008).

Figure 1. Effect of a number of factors on healing and wound closure in chronic wounds treated with decellularized, dehydrated human, amniotic membrane. Patients with more than 1 factor were less likely to achieve wound closure and the probability of wound closure was inversely related to the number of factors—whether or not good wound care was evidenced. The number of factors was a significant determinant of complete wound closure (P= 0.0027).ITT: Intent-to-Treat population; GWC: Good Wound Care population.

Several factors were compared individually between closed and unclosed wounds to determine significance of effect on closure (Table 6). As expected and consistent with the literature,¹⁹ the presence of infection prior to and during participation in the study was found to be a key determinant of wound closure. Figure 2 shows outcomes in wounds with infection prior to or during treatment with DDHAM are affected with or without other good wound care efforts. Infection was a highly significant determinant for closure of chronic wounds (P = 0.0002).

Figure 2. Effect of infection on wound closure in chronic wounds treated with decellularized, dehydrated human amniotic membrane (DDHAM). Infection prior to or during treatment of chronic wounds with DDHAM affected the outcome of complete wound closure with or without other good wound care efforts. Infection was a highly significant factor affecting chronic wound closure (P= 0.0002).ITT: Intent-to-Treat population; GWC: Good Wound Care population

The presence or absence of compliance with adequate offloading (for pressure ulcers and  diabetic foot ulcers) or compression (for venous leg ulcers) had a statistically significant impact on wound closure (P = 0.0131) in this observational study. There was no effect on the probability of achieving complete wound closure on the basis of the presence of peripheral arterial disease or venous hypertension/lymphedema (P = 0.07467) in this study group (P = 0.54767), although the extent or severity of the disease was not quantified for each case (Table 6). There were no serious or unexpected adverse effects related to DDHAM reported, and subject and investigator opinions were generally positive.

In the GWC group, the observation time for wounds that closed was a mean of 7.4 and median of 6.3 weeks (Table 7). The length of participation of compliant subjects with wounds that did not reach complete wound closure was a mean of 10.1 weeks and median of 8.9 weeks (P = 0.0027). The mean times for the ITT group were 8.0 weeks for closed wounds and 9.7 weeks for unclosed wounds. The 1.7 weeks was a highly suggestive difference although it did not meet the criterion for statistical significance (P = 0.0756, Table 4).

Thirty-two subjects with a variety of chronic ulcer types (venous, n = 14 [13 wounds]; diabetic foot, n = 10; pressure, n = 1; arterial [ischemic], n = 7 [4 wounds]) had failed previous courses of therapy with 1 or more advanced biologic therapies (ie, Apligraf, Organogenesis, Canton, MA; Dermagraft, Organogenesis, Canton, MA; Oasis, Smith and Nephew, Hull, UK; or Regranex, Smith and Nephew, Hull, UK). After a course of therapy that included the DDHAM allograft, nearly half (48.4%) of these ulcers closed despite previous biologic therapy failures. Those that did not close during a mean observation time of 10.3 weeks reduced in size from baseline by 50% (Table 8).

During the observation time of the GWC cohort the baseline area of the wound was of statistically significant importance (P = 0.0011). The rate of wound area reduction, however, did not differ significantly (n = 0.59) between the closed and unclosed wounds (Table 7).

In a prospective, randomized, controlled clinical study (PRCT) in a chronic wound population, such as patients with diabetic foot ulcers or venous stasis ulcers, patients would be excluded from the study with factors such as uncontrolled diabetes, significant peripheral vascular disease or severe venous insufficiency, autoimmune disorders, renal failure, concomitant steroid use, exposed tendon and/or bone, wound infection and/or osteomyelitis at the target wound site, Charcot foot, radiation treatment (previous or concurrent) of the target limb, active or recent history of malignancies or neoplasms at or near the target treatment site, significant malnourishment, or compromised immune status. The rationale for excluding patients with these coexisting conditions is that these factors are widely accepted as negatively impacting the patient’s ability to heal, and wound closure would require a longer period if it was at all possible. Furthermore, limiting the extent of variability within the study population allows for a more homogeneous study population in which the treatment effect is less diluted. With less heterogeneity, a smaller sample size is possible than what would be required if no limits on variability are placed. When looking at the criteria considerations listed above, in this DDHAM Use Registry Study, approximately 52% of the patients would have been excluded from a PRCT. Furthermore, some PRCTs further exclude patients based on baseline wound size and patient age. Twenty-two subjects were 80 years of age or over, with 10 subjects over the age of 85 years. Fifteen patients (16 wounds) had baseline wounds greater than 25 cm². It has been shown that the typical series of exclusion criteria in clinical trials in chronic wounds excludes more than 50% of potential participants.¹⁶ Therefore, applicability of PRCT outcomes to the expectations of performance in real-world wound populations raises questions about generalizing their results.²¹

The importance of offloading of pressure ulcers^22,23 and diabetic foot ulcers^24,25 as well as compression of legs with venous stasis ulcers^26,27 with regard to healing trends is supported by reports of outcomes in the medical literature. Consistent with these reported findings, in this observational study the presence or absence of adequate offloading (for pressure ulcers, diabetic foot ulcers) or compression (for venous leg ulcers) had a statistically significant impact on wound closure.

It is apparent that with the addition of each factor that potentially affects healing, the likelihood of wound closure in a similar timeframe decreases (average observation time of 8-10 weeks in chronic wounds). Good wound care (defined here as the absence of gross noncompliance with prescribed good wound care practices) improves the likelihood of wound closure. Control of these factors by limiting these characteristics in a study population via inclusion and exclusion criteria, and conducting a strictly monitored protocol of care, should permit a shortening of the treatment observation period with respect to the expectation of complete wound closure. Furthermore, attention to how data is captured along with monitoring the completeness and accuracy of data entered in collection tools will allow for more complete and accurate data listings and analyses.

In addition to analyzing factors influencing wound closure, the goal of the DDHAM Use Registry Study was to collect information on the clinical utility of a sterile DDHAM when used in management of a variety of chronic wounds. The authors found that a large number of subjects (n = 60, 49.6%), many with complicated wounds and comorbidities, reached complete closure of their wounds within an average of 8 weeks. The baseline size of the wound was the main, statistically significant factor driving these outcomes, which is in line with previous study findings.²⁸

Margolis and colleagues²⁸ combined patient-level data from the standard of care arms (ie, debridement, offloading, and appropriate wound dressings) of 5 diabetic foot ulcer PRCTs, with more than 500 subjects in total. Using strict inclusion and exclusion criteria the authors reported that, of diabetic foot ulcers with a mean baseline wound area of 1.6 cm², 23.9% healed at 12 weeks and 32.8% at 20 weeks. They also noted that baseline size was a key determinant for healing in a specific treatment period of time. While the wound size distribution of the wounds in that article cannot be compared with the distribution in the group of diabetic foot ulcers in the current study, it is noteworthy that, in the authors’ analysis, the mean wound area at baseline of the ulcers that continued to wound closure—within 8 weeks on average—was substantially larger (3.1 cm²).

Additionally, in another chronic wound type, Mostow et al²⁹ demonstrated that 34% (20/58) of the control (standard care) arm in a venous leg ulcer study closed at 12 weeks with compression dressings and debridement as a standard of care. Also, in Falanga and Sabolinski’s pivotal venous ulcer study,³⁰ the control arm (n = 100) achieved an incidence of complete closure in approximately 24% of the ulcers at 12 weeks.

The rates of wound area reduction between closed and unclosed wounds in the current study were similar, but baseline wound sizes were significantly bigger in those that did not close in similar observation times. Given the significantly larger wound size in the unclosed population and the relatively short treatment observation period (about 8 weeks; Tables 6 and 7), additional treatment observation time might have allowed for a higher percentage of complete wound closure.

The study did not have a control arm nor were the wound measurements provided at regular intervals, but done only on enrollment and at the end of the study. Therefore, it is difficult to determine how the application of DDHAM compares to other treatment regimens with respect to wound closure outcomes. In addition, the average period during which subjects participated was shorter than in most studies. This, again, makes comparisons to other study regimens difficult.

This study demonstrates what can be expected when treating chronic wounds in patients with multiple comorbidities in the real-world setting, without the restrictions of inclusion and exclusion criteria as they are used in a PRCT. The chronic wounds population in this study was gathered under 1 inclusion criterion (ie, presence of a uninfected chronic wound that may benefit) and 2 exclusion criteria (ie, known hypersensitivity to the DDHAM and/or infection present upon enrollment), leading to a heterogeneous population.

Similar to other studies, the authors found the presence of comorbidities has an influence on wound closure. The presence of infection and noncompliance with appropriate pressure-offloading or compression therapy necessary as adjunctive treatments for underlying disease, related to wound etiology or comorbidity, also influences the incidence of wound closure. The impact on complete closure rates by baseline wound size is determined by the length of the observation time. Larger wounds take longer to heal than smaller ones when the rates of area reduction are similar.

This study also demonstrated the clinical benefit of DDHAM treatment: the chronic wound GWC population study demonstrated an incidence of complete wound closure of about 50% at a mean time to closure of approximately 8 weeks. This higher rate of wound closure, compared with what was reported for control groups in other chronic wound PRCTs (24%-34%)^29-32 most likely is related to the use of DDHAM in combination with the standard of care in 2005. This is a demonstration of effectiveness in a broad, real-world population of all types of chronic wounds. In the health care environment of 2015, where quality metrics and real-world data are being sought, this registry is an example of the value of real-world data in the arena where health care providers and patients are dealing with the challenges brought by the complexities of chronic wounds.

There was no significant difference in the rate of reduction of wound area (P > 0.5) in the closed and unclosed population. Given the larger wound size in the unclosed population and the relatively short treatment observation period (about 8 weeks), additional treatment observation time might have shown an increase in complete closure. Better control of factors that negatively impact wound closure through a monitored protocol with stricter inclusion and exclusion criteria also might have shown higher closure rates, but that approach would have diminished the impact of a real-world study. Real-world studies may help discern which, if any, comorbidities or other potentially interfering factors impede the healing process. However, this type of observation without applied limitations for enrollment potentially dilutes the extent of complete wound closure that is traditionally achieved over set time limits in strictly controlled and monitored studies (ie, PRCTs).

Validation of the clinical effectiveness of a regimen that includes DDHAM in a PRCT in chronic wounds with stricter enrollment criteria and monitoring of a standard of good wound care is recommended.

Janice M. Smeill, MD, and Helen D. Hahn, RN, MBA are from Alliqua Biomedical, Langhorne, PA; Terry Treadwell, MD is from the Institute for Advanced Wound Care, Montgomery, AL; and Michel H. Hermans, MD is from Hermans Consulting Inc, Newtown, PA.

Address correspondence to:

Janice M. Smiell, MD
Alliqua Biomedical
jsmiell@alliqua.com

Disclosure: Janice M. Smiell and Helen D. Hahn are employees of Alliqua Biomedical. Dr. Smiell has a finanical interest in Celgene Corp, Summit NJ. Terry Treadwell is a consultant for Alliqua Biomedical and the Study Investigator; as clinical editor of Wounds, he had no involvement in the peer-review of this manuscript.