Clostridial Collagenase for the Management of Diabetic Foot Ulcers: Results of Four Randomized Controlled Trials

Background. Despite major treatment advances, diabetic foot ulcers (DFUs) remain a frequent and debilitating complication of diabetes mellitus and a major cause of significant morbidity and mortality. Objective. This study evaluates and compares clinical effectiveness of clostridial collagenase ointment (CCO) with standard care (SC) in patients with DFUs, with the goal to define best clinical criteria under which to use CCO in the DFU population. Materials and Methods. This is a pooled data analysis of 4 randomized controlled trials that compared clinical effectiveness of CCO to SC. A total of 174 adult patients with 1 target DFU each who underwent treatment with CCO (n = 88) or SC (n = 86) for 4 or 6 weeks were evaluated. Assessments included wound area reduction, wound bed status, and time to closure. Results. Statistically significant mean percentage change in wound area from baseline was numerically greater for CCO than SC at the end of study (EOS) following 6 or 8 weeks of treatment; these values were -56% and -10%, respectively, in the subgroup of plantar surface ulcers (P = .05) and wounds assessed as “low necrosis” (≤ 25% necrotic) at baseline (-64% vs. -20%). When rapidly healing ulcers were excluded from the analysis, the difference in ulcer area reduction was even greater for CCO compared with SC at EOS (-53% vs. -7%; P = .05). Conclusions. Active CCO therapy was associated with a nonstatistically greater reduction in wound size than any of the passive or mechanical SC modalities at end of treatment. This was statistically significant when used in conjunction with sharp debridement and for slow healing ulcers, larger sized wounds, or plantar surface wounds.

Diabetic foot ulcers (DFUs) are a frequent and debilitating complication of diabetes mellitus and a major cause of significant morbidity and mortality. Persistent nonhealing DFUs have a significant negative impact on health-related quality of life.¹ Diabetic foot ulcers carry an increasing risk for major amputations for each year they remain open and account for more than two-thirds of all nontraumatic amputations performed in the United States.^2-4 Standard care (SC) of DFUs includes offloading with specialized shoes and adaptive equipment (eg, crutches) to minimize pressure on the wound, systemic and/or topical antimicrobials to decrease the bacterial burden in the wound and when indicated for infection, and local measures to promote an optimal wound environment for healing.^5,6 Effective local measures include the use of occlusive dressings, eg, foams and polyurethane films,^7,8 to maintain a moist environment as well as debridement. Much of the underlying strategy guiding wound management is based on the concept that a clean wound, with minimal exudate and a completely granulated wound bed (similar to a wound ready for skin grafting), is likely to heal. Preparation of the wound bed to achieve such a state is a critical element of DFU management.^9,10 

Debridement techniques used in clinical practice for wound bed preparation vary from passive moist dressings (“autolysis”) to active surgical, enzymatic, and mechanical debridement.^11-13 Clostridial collagenase ointment (CCO; Collagenase Santyl Ointment; Smith & Nephew, London, UK) is the only enzymatic agent approved by the United States Food and Drug Administration (FDA) for debridement of wounds and burns. It has been shown to specifically and preferentially digest native collagens without harming healthy tissue and effectively removing nonviable debris.¹⁴ It is well known that substrate specificity of a protease depends upon the peptide sequence configurations for binding to the active site of an enzyme. As a protease increases in specificity, it is more able to target only a few specific sequence configurations. Collagenase is one such protease, and it is capable of causing hydrolytic cleavage specifically in the triple-helical region of collagen molecules.¹⁵ Hu et al¹⁵ evaluated the relative substrate activities (kcat/Km) of an immobilized heptapeptide library with the class I Clostridium histolyticum collagenase and found that only a few sequences displayed high specificity with this enzyme. Proteases with broader substrate spectrums, such as papain and trypsin, cleave multiple peptide bonds and degrade a much more nonspecific and potentially more beneficial group of proteins.¹⁶ The selectivity of Clostridium collagenase and papain have been assessed for their proteolytic activities on collagen (type I) and elastin.¹⁶ Collagenase displays much higher collagenolytic activity and much lower elastinolytic activity than papain. This increase in specificity decreases the potential harm to surrounding tissue. Although debridement of wounds is thought to be a key element in wound bed preparation that leads to improved healing,¹⁷ there is a paucity of clinical data from well-controlled trials to confirm this widespread belief or shed light on what forms of debridement are most effective.¹⁸ It should be noted that debride means “unleash,” therefore the concept of removing devitalized tissue is one way to help unleash the ability of the wound to heal. However, it is not the sole means; it is a mistake to equate debridement exclusively with removal of visible debris. Less visible but no less important factors such as biofilms or senescent cells must also be considered. 

The present study is an analysis of 4 small, prospective, randomized controlled trials (RCTs) in which DFUs were treated either with CCO or various SC methods. Together they represent the largest prospective study of CCO published to date. The goal of this analysis is to define the best clinical criteria under which to use CCO in the DFU population. These exploratory RCTs were designed to provide descriptive data and generate testable hypotheses relevant to the clinical outcomes associated with CCO debridement of a DFU. Importantly, these trials have very similar patient populations and designs and have been analyzed both independently and concordantly as pooled groups, with subgroup analyses employed to explore response to treatment as influenced by characteristics of the ulcers such as location, size, and appearance.

Ethics statement
The trial protocol, investigators, and consent documents for each clinical trial were reviewed and approved by accredited Institutional Review Boards. The 4 studies were conducted in accordance with the ethical principles originating in the Declaration of Helsinki and with the applicable regulatory requirements in the Good Clinical Practice guidelines by the International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH). All patients participating in the 4 clinical trials provided written informed consent. The 4 clinical trials evaluated in this study are registered at ClinicalTrials.gov: NCT01143714; NCT01143727; NCT01408277; and NCT01056198.

Trial participants
In brief, eligible patients were 18 years or older and diagnosed with type 1 or type 2 diabetes and a neuropathic nonischemic foot ulcer of at least 1 month’s duration between 0.5 cm² to 10 cm² in area, with the exception of trial 2 (NCT01143727) (area > 1.5 cm²; no upper limit). Wound size for eligibility into the study was determined at the screening visit. Baseline visits and start of treatment were about 1 week later to allow for the results of eligibility labs. During this time period, some subjects’ wounds decreased in size while others increased, causing the wound area to range in size from 0.1 cm² to 24 cm² at the start of treatment.

Trial 1 (NCT01143714)
The primary objective of this randomized, double-blind, CCO versus vehicle-controlled trial was to assess the percentage of wound area change from baseline at the end of 4-week treatment and 12-week trial periods. Secondary objectives were proportion of closed wounds within 12 weeks, rate of wound area reduction over the 4-week treatment phase, periwound skin status, and incidence of wound infection. For this trial, the test articles were supplied in prenumbered kits given in sequential order to eligible subjects as they entered the study. Study personnel and subjects remained blinded as to treatment.

Trial 2 (NCT01143727)
This randomized, open-label trial was a comparison of CCO and hydrogel on mildly inflamed DFUs (defined as the presence of 2 or more of the following symptoms: purulence, erythema, pain, tenderness, warmth, or indurations). The primary objective of the trial was to describe the weekly wound appearance assessed by a standardized wound assessment tool. The secondary objectives were to assess the percentage of wound area change from baseline and measurement of wound fluid levels of endogenous proteolytic enzymes and inflammatory factors.¹⁸

The study site was provided with a set of envelopes, numbered in sequential order, to randomize eligible subjects into 1 of the 2 treatment groups (CCO or hydrogel). An envelope was only opened, revealing the treatment assignment, once the baseline measurements had been taken for an eligible subject. The eligible subject was assigned the randomization number printed on the envelope. Within each envelope was a card containing the identification of the treatment to be given to the subject; the contents of the envelopes had previously been determined by a randomization sequence. All site personnel and subjects were aware of the test article received by each subject.

Trial 3 (NCT01408277)
This was an open-label trial in which patients were randomized to debridement with either CCO applied daily (after a single sharp debridement at baseline) or mechanical debridement using saline-moistened gauze (SMG) plus intermittent elective sharp debridement. The primary objective of the trial was to describe changes in the wound status total score using a standardized wound assessment tool (a modified Bates-Jensen Wound Assessment Tool [BWAT]) and 8 subscale scores at the end of the 4-week treatment and 12-week trial periods, while the secondary objective was the percentage of wound area change from baseline¹⁹ at the same time points. 

For this study, central randomization across all sites was used to assign the test article. Once a subject was eligible, the site called a central number to receive the next sequential test article for that subject according to the schedule held by the randomization center. Both the clinician and the patient were aware of the test article received. 

Trial 4 (NCT01056198)
This was a randomized, parallel group, open-label trial comparing CCO, used as an adjunct to sharp surgical debridement, to sharp surgical debridement with various SC regimens (at the discretion of the principal investigator). The primary objective was to assess the percent of wound area change from baseline at the completion of the 6-week and 12-week trial periods. The secondary objectives were assessment of the wound status, measured using the standardized wound assessment tool including 8 subscale scores, and the number of sharp debridements required during the trial period, measured at the same time points. The time to wound closure for each group was also assessed.^20,21

For this study, central randomization across all sites was used to assign the test article. Subjects qualifying for enrollment were issued a randomization number and assigned to treatment by the electronic data capture system, which had access to the randomization code.

Trial design
All trials were conducted at a mixed panel of academic wound care programs, outpatient wound care facilities, or private practices located in the United States. All study visits, procedures, and data collection were performed at the individual study sites. Patients were randomly assigned to treatment (allocation ratio, 1:1) to prevent treatment allocation bias. The randomization sequences for each study were generated by a consultant biostatistician in blocks of 2, using the SAS Version 9.1.3 random number generator (SAS Institute Inc, Cary, NC). The subjects at each site were enrolled by the Investigator and the site study coordinator, with treatment assignments made according to the procedures outlined in the appropriate protocol (see specific trial above). 

eFigure 1 illustrates the overall trial design of each trial included in this report. A double-blind design was used for Trial 1 but not in the other 3 due to the obvious dissimilarities in the treatments. With the exception of Trial 2, each trial entailed a screening visit followed by a 5- to 9-day screening period before randomization to treatment for 4 weeks (Trials 1–3) or 6 weeks (Trial 4) with a follow-up interval terminating 12 weeks after randomization. Trial 2 had the same design but did not have a follow-up period (eFigure 1). In general, all studies incorporated some component of autolytic debridement in the SC cohort, whether this was a hydrogel or a moisture-retentive primary layer such as silver sulfadiazine. 

Wound care during the 5- to 9-day screening interval was at the discretion of the local Investigator with certain restrictions (no CCO, topical antibiotics, or engineered tissue constructs). Dressings were changed once daily throughout trial participation (screening, treatment, and follow-up). In patients assigned to CCO, a 2-mm thick layer of ointment was applied to the wound surface once daily. Direct contact SC regimens employed for the control groups in all trials were also applied once daily. The specific topical therapy for SC was petrolatum in Trial 1, hydrogel (Tegaderm; 3M, St Paul, MN) in Trial 2, and SMG in Trial 3. In Trial 4, SC was selected at the discretion of the investigators and included hydrogel, SMG, alginate, or silver-impregnated dressings. 

During the posttreatment follow-up period of Trials 1, 3, and 4, subjects whose wounds remained open (following either CCO or SC treatment) were treated in parallel fashion. For Trials 1 and 3, a silicone screen (Mepitel; Mölnlycke Health Care, Norcross, GA) was used to cover the wound bed and secured with dry gauze. Allevyn foam (Smith & Nephew) secured with dry gauze and Coban (3M) was employed in Trial 4. All of these therapies obviously incorporate a component on autolytic debridement that, although not standardized for the outlined reasons, does represent autolytic debridement nonetheless. 

Offloading and debridement
In all trials, patients were required to use offloading shoes and were provided with a Darco MedSurg (Darco International, Huntington, WV) padded shoe with an accommodative insert containing removable pegs (removal of pegs was discretionary). Debridement of the wounds with sharp instruments, such as scalpels and curettes, was governed by protocol-specific rules that differed by trial. The same rule was applied to both treatment groups in each trial except Trial 3 where sharp debridement was not permitted post baseline for patients assigned to CCO unless medically required and, if performed, the CCO patient exited the trial. Sharp debridement was performed in Trial 4 if any undermining, necrotic tissue type, necrotic tissue amount, exudate type, or exudate amount scored ≥ 3 on the subscales of the modified BWAT at baseline or any subsequent visit.

Wound measurements
Digital cameras coupled with laser imaging devices (ARANZ Medical Ltd, Christchurch, New Zealand) were used to measure wound area for Trials 1, 2, and 4. Wound area was measured manually (length x width) for Trial 3. The criteria for the target wound to be considered as closed included full (100%) epithelialization, no drainage, and no requirement for a dressing. Wound bed assessment was performed using an 8-category tool modified from BWAT²² for Trials 2–4; wound bed of Trial 1 was not assessed using any numeric tool. The wound status total score incorporates various factors ascertained before debridement if debridement was performed. The factors scored individually included wound edge appearance, undermining, necrotic tissue type, necrotic tissue amount, exudate type, exudate amount, periwound skin color, and granulation tissue amount.

Pooled data assessments
Since all 4 randomized trials had similar treatment designs (eFigure 1) with comparable inclusion and exclusion criteria (eTable 1), data were pooled to assess the overall clinical effectiveness of CCO in comparison with SC in patients with DFUs. The variables assessed in the pooled data analysis included wound area and wound closure at the completion of 4 weeks of treatment as well as at end of study (EOS). In addition, subgroup analyses were performed based on wound size (above and below median size), plantar versus nonplantar surface wounds, and baseline necrosis and granulation tissue wound assessment tool subscores.

Statistical procedures
The purpose of these pilot studies was to provide descriptive data regarding the outcomes achieved following 4 to 6 weeks of treatment with CCO used as indicated for the debridement of chronic DFUs. Because the intent of the studies was to generate rather than to test hypotheses, sample size was not based on statistical power calculations. 

The intent-to-treat (ITT) population consisting of all patients randomized to treatment was the source of pooled data for all analyses. Data analysis was performed using SAS Version 9.1.3 (SAS Institute Inc). Summary statistics including distributions for single variables are provided for baseline data. Wound sizes at the randomization visit for each trial were defined as the baseline. The average percentage change from baseline in the wound area and status for the 2 treatment groups were compared at the end of treatment (EOT) (4 weeks in Trials 1–3 and 6 weeks in Trial 4) as well as at the EOS (Trials 1, 3, and 4 only) using a mixed-effects analysis of covariance (ANCOVA). Treatment and treatment week were defined as fixed effects and patient as a random effect. The corresponding wound area at baseline was used as a covariate. Missing values for any of the individual subscales were imputed with the mean score for that assessment (at that visit) or using the method of last observation carried forward as defined a priori in the respective protocols. Student’s t-test was used to compare the average percentage change from baseline in wound area, and Fisher’s exact test was used to compare the differences in proportions of wounds healed.

As noted above, these studies were planned as pilot studies to provide guidance to future studies with CCO and to define the best clinical criteria under which to use CCO in the treatment of DFUs. There was no expectation that data from the individual studies would provide significant statistical information.

In this report, 177 patients with DFUs were randomized in the 4 evaluated trials. Of these subjects, 174 were evaluable and included in the ITT population for analysis (eTable 2). Demographic parameters and wound characters are shown for the individual trials and for the pooled analysis of patients in eTables 3 and 4, respectively. The demographic characteristics of patients in the 4 trials were comparable in terms of age, gender, ethnicity, and race. Trials 1, 3, and 4 had comparable mean and median wounds sizes (mean, 2.2 cm²), but Trial 2 had significantly larger wounds at a mean size of 8 cm², a consequence of no upper size limit in Trial 2 which required “mildly inflamed” ulcers at baseline. 

All analyses were performed using the ITT population, which consisted of the CCO treatment group (n = 88) and the SC treatment group (n = 86) (eTable 2).

The trajectory of wound closure, defined as the percentage change from baseline in wound surface area, is depicted by week for the pooled CCO and SC groups in eFigure 2. At 4 weeks, the reduction in area for CCO-treated wounds was 43% versus 19% in the combined SC wounds (not significant; NS). The corresponding area change when EOT is used for comparison (6-week data from Trial 4 substituted for 4-week data) was 52% for CCO and 25% for SC (NS).  At EOS, the corresponding results were 55% with CCO and 25% with SC (P = .06). 

When the same area analysis was restricted to the plantar wound subgroup, the results shown in eFigure 3 were obtained. The pooled analysis of all plantar wounds found that at EOT the reduction in area was 60% for CCO and 36% for SC (NS). The corresponding difference in area reduction at EOS between the 2 groups was statistically significant: 56% for CCO and 10% for SC (P = .0497). Similarly, when data were analyzed by individual trial, the reduction in wound size was greater, but not statistically significant, for CCO than SC at both EOT and EOS in each trial. 

The influence of baseline wound size on the response to therapy was examined through subgroup analysis of wounds larger or smaller than the median wound size. As shown in eFigure 4, when all wounds are compared based on treatment group, the reduction in size compared with baseline at EOT was 43% for CCO and 19% for SC. The corresponding changes at EOT in ulcers less than the median area of 1.6 cm² (mean area, 0.83 ± 0.38 cm²) was 60% for CCO and 49% for SC. When ulcers ≥ 1.6 cm² (mean area, 4.3 ± 4.0 cm²) were assessed at EOT, the CCO-debrided wounds decreased 48% and the SC wounds decreased 15% in size. The EOS measurements showed an average decrease for all CCO-debrided wounds of 57% versus 19% for SC. The corresponding change in size for smaller wounds was 78% for CCO versus 52% for SC; for larger wounds, CCO treatment led to an average decrease of 48% compared with an increase in area with SC of 19%. Data show that independent of treatment, small wounds tended to experience greater mean decreases in area than large wounds. This is accounted for in part by the rate of keratinocyte migration. If keratinocytes can be assumed to migrate at the same rate independent of ulcer size, small wounds will have larger percentage decreases over the same time period than larger wounds. Nevertheless, the degree of wound closure associated with CCO compared with SC was more pronounced for large wounds at both EOT and EOS with the difference approaching significance at EOS (P = .10). 

When the influence of wound location was examined, complete healing in both treatment groups was similar for nonplantar wounds compared with wounds located on the plantar surface. At EOT, 13% (19/149) of all plantar wounds (independent of treatment arm) were healed compared with 16% (3/19) of all nonplantar wounds (P = NS). However, the pooled data for Trials 1, 3, and 4 at EOS showed 30% (45/149) of the plantar wounds were healed (independent of treatment arm) compared with 63% (12/19) of nonplantar wounds (P = .10).

Similar to wound location (independent of treatment arm), wound size also appeared to influence the likelihood of complete healing. By EOS, 46% (37/80) of all small wounds (< 1.6 cm²) achieved complete closure compared with 23% (22/94) of all large wounds (≥ 1.6 cm²); the difference in proportions was highly significant (P < .01). Median time to closure for wounds achieving closure was 10 weeks for all CCO-debrided wounds compared with 12 weeks for nonenzymatically treated wounds. Median time to closure for both larger and smaller wounds debrided with CCO was 10 weeks compared with 11 weeks for smaller wounds and 12 weeks for larger wounds treated with SC. 

Assuming DFUs that heal within 4 weeks of initiating treatment represent a class of wounds less likely to be influenced by the details of the local care regimen employed (perhaps because of better patient compliance with offloading or other factors) and wounds that fail to heal by 4 weeks represent a more refractory class of wound,²³ the reduction in wound area was examined when wounds  that healed by week 4 were excluded from the pooled analysis. As shown in eFigure 5, the reduction in wound area at EOT was 47% for CCO debridement and 22% for SC (P = .23); at EOS, the corresponding reductions were 53% and 7%, respectively (P = .05). 

The association between reduction in wound area and degree of granulation present at baseline was examined for Trials 2–4. For relatively well-granulated wounds (≥ 75% granulation), there was a moderately greater proportion of closures associated with CCO at both weeks 4 and 12 (EOS) compared with SC. Mean percentage reduction in wound area at 4 weeks was 46% with CCO versus 38% with SC (NS); the corresponding results at 12 weeks were 60% versus 21%, respectively (NS). The data for less granulated wounds (< 75% granulation) showed greater numerical differences in favor of CCO: mean reduction in area with CCO was 71% versus 35% for SC (NS) and 65% versus 8% at EOS (NS). Similar results were obtained at EOT when 6-week data from Trial 4 were substituted for 4-week measurements (NS).

A similar analysis was performed for those wounds with low to moderate levels of necrosis at baseline (subscale score ≤ 2; < 25% of surface necrotic; 85/120 ulcers [71%]). Debridement with CCO was associated with greater average decrease in wound area for low to moderate necrotic wounds compared with SC; 4-week average percentage reduction in area was 67% with CCO versus 23% with SC and was 64% versus 20% at 12 weeks (NS). 

Stalled wounds, defined here as wounds which failed to achieve > 10% closure compared with baseline, were compared within the pooled CCO or SC therapy groups. As shown in eTable 5, the number of stalled wounds in the pooled data was greater for SC than CCO at both EOT (23% vs. 11%; P = .04) and at EOS (23% v. 15%; NS). 

The clinical use of Clostridium histolyticum-derived collagenases, proteases, and peptidases as a method to achieve ulcer and burn debridement was described by Howes et al²⁴ in 1955, following more than a decade of work on the isolation and characterization of various Clostridium enzymes. A thorough review of enzymatic debridement by Sherry and Fleteher²⁵ highlighted the challenges to overcome as of 1960, with CCO being a newer idea under consideration. Subsequently,  Altman et al²⁶ reported 1 of several organized clinical evaluations of collagenase for the debridement of DFUs.Since that time, there have been various postulates about how to best use collagenase in the diabetic foot. However, based upon a PubMed key word search, there has not been a definitive study which demonstrates best practices until now. Therefore, the results from these studies should be looked at as a guide on how to achieve best results when using CCO on the DFU.

The 4 pilot studies reported herein were not planned to give statistically significant results, but rather to provide guidance to future studies on CCO and to define the best clinical criteria under which to use CCO in the treatment of DFUs. With 174 patients, the present authors report a large prospective evaluation of this topical enzymatic debrider as it affects DFU progression.

The efficacy of CCO in the debridement of chronic ulcers has been established through multiple clinical trials.^27-29 Efficacy of Clostridium collagenase in the diabetic foot stems from its unique properties. This bacterial enzyme can degrade the triple helix of human collagen,³⁰ attacking it at several points along the polypeptide chain, in contrast to endogenous collagenases. The chronic DFU is noted to have a base of necrotic tissue with cellular debris embedded in an extracellular matrix composed of types I and III collagen, glycoproteins, and proteoglycans. Evidence suggests that debridement with CCO may render the wound bed permissive for keratinocyte migration or, via the release of peptide fragments, stimulate proliferation and migration of keratinocytes and fibroblasts.³¹

The wound assessment tool used in 3 of the 4 trials discussed herein was based on 8 subscales (edges, undermining, necrotic tissue type and amount, exudate type and amount, skin color around wound, granulation) of the modified BWAT. However, many of these fields are not particularly germane to the effect of topical enzymatic debridement in the DFU. Therefore, the authors have focused on granulation and the amount of necrotic tissue as these appear to be the variables that may help the clinician decide which wounds may respond best to topical exogenous collagenase.

In these trials, the baseline necrotic score was recorded prior to debridement when debridement was allowed. For 9 wounds that had a subscale score of ≤ 2 (which correlated to < 25% of the wound bed covered with necrotic material) versus ≥ 3 (which correlated to > 25% of the wound bed covered), the CCO group outperformed the SC group. When the necrotic score was ≥ 3 and sharp debridement was limited or not allowed in the CCO group but allowed without limitation for SC, the CCO group did not perform as well, emphasizing the particular benefit of adjunctive baseline sharp debridement in ulcers with substantial amounts of necrotic tissue. It also may speak to the fact that no matter the treatment, wounds with excessive slough behave poorly. 

Adverse events, both volunteered and elicited, were monitored across all 4 trials. None of the reported adverse events were considered to be related to the test article. The safety profile of the subjects treated with CCO was very similar to that reported by the subjects treated with SC.

Overall, the pooled data analysis showed CCO debridement to be associated with greater reduction in wound size than that seen with SC methods in all subgroups (plantar location, small wound size, large wound size, wounds remaining open after 30 days, those with low to moderate levels of necrosis at baseline, and those with high levels of granulation). One inference is the authors’ preference to pair CCO therapy with initial and repeated sharp debridement when significant amounts of necrosis are present is justified. Further, most clinicians would agree that over a short time period such as 1 month, CCO debridement does not obviate the need for sharp debridement when high levels of grossly necrotic tissue are present. Moreover, with longer periods of debridement, as evidenced in the Milne studies,^29,32 CCO can remove significant amounts of nonviable tissue and accomplish complete debridement without concomitant sharp procedures. These inferences are supported by the results of Trial 4, where the longer period (6 weeks) of CCO usage and the adjunctive use of sharp debridement were associated with better outcomes for the CCO cohort than seen in Trial 3, which only allowed for 4 weeks of enzymatic debridement and no sharp debridement. The authors strongly believe these disparities highlight the need for standardized debridement guidelines that are very strictly defined. While the Centers for Medicare and Medicaid Services continue to better define appropriate levels of debridement, the limited number of extant prospective trials describing practice patterns differ substantially. Numerous prospective DFU studies try to mandate standardized debridement techniques through proctoring, videos, or courses (authors’ personal experience); the presented trials did not. In addition, grading of the amount of necrotic tissue was obtained by best visual estimate. While beyond the scope of this article, further definition, standardization, and implementation of improved debriding algorithms are necessary. 

As noted elsewhere, there is consistent support in the literature for some amount of sharp debridement, especially initial debridement, but the ideal frequency has not been definitively elucidated.^18,32-36 While frequency of sharp debridement was addressed in the present trials, due to the protocol-mandated strictures on the use of sharp debridement, the data do not provide additional clarity other than what has already been defined.^18,33 In Trial 4, sharp debridement was employed almost weekly in all patients in both arms. This indicates that very good outcomes can occur when CCO is used in conjunction with frequent sharp debridement. While superior to SC (SMG plus sharp debridement), outcomes with CCO in Trial 3 (where sharp debridement was not allowed) were not as good as those seen when CCO was employed with sharp debridement (Trial 4). In Trial 4 where weekly debridement occurred, the wound area reduction at the end of 4 weeks was 2-fold better for CCO relative to SC, but was somewhat less when only the initial sharp debridement was allowed and wounds were also debrided with CCO.

The results presented herein demonstrate that CCO resulted in better progress toward healing in most of the subgroups studied, but perhaps the greatest benefit was seen for those wounds least likely to heal, ie, in recalcitrant wounds. Therefore, while CCO can be of benefit in the management of most DFUs, it should certainly be considered for use in DFUs that have fallen off a correct wound healing trajectory as defined by Zimny et al.³⁷

It should be noted that like most other DFU studies, many predictors of poor DFU outcomes, such as end stage renal disease, local infection, and significant peripheral vascular disease, were excluded from all trial cohorts in order to provide a more homogeneous group.³⁷ While the authors strongly believe the plantar location is associated with delayed healing in the DFU, the largest prospective trial of DFU healing predictors (Eurodiale)³⁸ did not show ulcer location (plantar/nonplantar) or depth to be predictors of nonhealing, though ulcer size was > 1 cm² and associated with an odds ratio (OR) of nonhealing at 1 year of 1.25 and > 5 cm² having a 3.48 OR of nonhealing. Other findings in that study associated with failure to heal were polyneuropathy, poor functional status, duration of the ulcer, and age of the patient.³⁸ It is possible that the longer follow-up period (52 weeks) in the Eurodiale trial is responsible for these differing outcomes.³⁸ Nevertheless, consistent with the results reported herein and with other trials, plantar wounds and larger wounds were found to close more slowly than smaller, nonplantar wounds.^38,39 It is evident from this data that these “anatomically and size challenged” wounds do better with CCO than SC. 

In closing, the trials included in the present analysis are similar enough to allow for the analysis of the pooled data. Interestingly, the SC treatment groups from each of the 4 trials worsened after the acute treatment phase (eFigures 2, 4), while the CCO treatment groups maintained or continued slight improvement during this period. This observation suggests possible salutary effects of CCO beyond the treatment period that are not present with the various SC used.

Obviously merging 4 heterogeneous trials that are too different to create a meta-analysis provides some limitations. However, the prospective, controlled, multicenter and adjudicated nature of the trials allows for some validity.  The fact that all study designs were not identical with just the control changed requires some assumptions as far as similarities and outcomes. In general, while these studies were designed with similarities, they were not designed to provide a statistically significant outcome. Also, since CCO is a debriding agent, it would have been ideal to have complete debridement as a primary outcome across all studies; however, adequacy of debridement remains an elusive study outcome, as adequate objective measures do not yet exist to measure this outcome.  In addition, the lack of standardization of debridement techniques may have increased the variability between groups. 

The conclusions derived from this large, prospective dataset were that CCO was effective when applied once daily for 4 to 6 weeks following an initial sharp debridement and better outcomes were obtained when coupled with weekly sharp debridement in cases where large amounts of necrotic debris were present at baseline. The benefits of enzymatic debridement with CCO were even more apparent in stalled or slowly progressing wounds, wounds > 2 cm², and those on the plantar surface of the foot. In addition, there is a sustained and ongoing reduction in the size of effectively debrided DFUs after cessation of therapy. Clostridial collagenase ointment appears to be an appropriate debridement modality for use in DFUs. These generalizable observations can serve as a basis for the appropriate inclusion of topical maintenance debridement guidelines.

Affiliations: Division Vascular/Endovascular Surgery, Department of Surgery, Mount Sinai St. Luke’s and Mount Sinai Roosevelt Hospitals, New York, NY; and Division of Vascular Surgery, Department of Surgery, University of California Irvine Medical Center, Orange, CA

Correspondence:
John C. Lantis, II, MD, FACS
1090 Amsterdam Avenue, Suite 7A
New York, NY 10025
JLantis@chpnet.org

Disclosure: Drs. Gordon and Lantis are on the speaker’s bureau for Smith & Nephew (Fort Worth, TX) and served as Principal Investigators in one of the clinical trials discussed in this article. They have no financial or proprietary interest in any material or method mentioned. These studies were funded by Smith & Nephew. G. Kesava Reddy, PhD, MHA, of Creative Medical Communications, LLC, and Renée Carstens, Smith & Nephew, provided medical writing support.