The Effect of a Biofilm-disrupting Wound Gel vs. a Broad-spectrum Antimicrobial Ointment on a Chronic Wound Microbiome: A Secondary Analysis Associating Clinical and Laboratory Findings

The incidence of chronic wounds is a growing health care problem. According to Armstrong et al, someone develops a DFU every 1.2 seconds,¹ with amputation occurring every 20 seconds.² Chronic wounds, including DFUs, venous leg ulcers, and pressure injuries, affect over 8.2 million Medicare recipients.³ In developed countries, 1% to 2% of the population will experience a chronic wound in their lifetime.^4,5 Not only are nonhealing wounds catastrophic to patients and their families, but Sen³ reported estimated calendar year 2014 health care cost projections ranging from $28.1 billion to $96.8 billion, describing surgical wounds and DFUs as the costliest to treat.

A highly colonized wound can support bacterial loads that do not appear infected but can interrupt the healing process. Recognizing that bacterial load may influence healing, Kim et al⁶ questioned whether reducing wound bioburden would affect wound size over time and if it could be considered a predictor of healing. Sophisticated techniques to assess bioburden, such as PCR, are becoming more readily available and standard of care.⁷ The original study samples “were analyzed in 2 ways: (1) by quantitative real-time PCR test for bacteria and fungi (which also included a qualitative real-time PCR test for resistance factors to vancomycin and methicillin); and (2) by DecodEX Microbial Genetic Identification Sequencing (MicroGen DX [hereafter, proprietary genetic sequencing]) to detect bacterial organisms and fungal pathogens that may be present in patient specimens.”⁶ The effectiveness of combining topical treatments (experimental [novel biofilm-disrupting wound gel] and control [triple-antibiotic maximum-strength ointment]) designed to reduce bioburden with sharp debridement was measured at the first visit and after 1 month. In that study, wound size in the experimental group decreased significantly, with a 71% reduction in wound area, compared with 24% for the control (P <.001). Additionally, by 12 weeks 53% of patients in the experimental group achieved closure, compared with 17% in the control group (P <.01).

In 2021, experienced wound providers met to discuss best practices for wound care.⁷ Several aspects of wound care were evaluated. After considering previous studies, the panel agreed that all chronic wounds have contamination or infection, requiring management of wound bioburden. There was also agreement that if after 4 weeks of standard care the wound area is reduced by 50%, then it is likely to heal by 12 weeks, based on previous research.^7-10 Historically, sharp debridement was identified as a leading method of affecting the wound microbiome; however, in a study by Verbanic et al,¹¹ the association of wound microbial levels and reduction of bioburden from debridement alone did not directly alter the total microbial bioburden or wound closure. However, other research indicates a link between levels of bioburden and wound healing.^12,13 With growing interest in understanding how bioburden affects healing, more treatment methods are needed to optimize the local wound environment to reduce bacterial bioburden and improve wound healing.¹⁴

The present secondary analysis explores the effects of bioburden levels on wound size reduction and wound closure.

Inclusion criteria and treatment

The original study was a 12- to 16-week, 2-site, prospective, randomized, open-label clinical study of patients with recalcitrant chronic wounds performed from September 2014 through March 2016.⁶ Both the Mayo Clinic Institutional Review Board (Rochester, MN) and the Schulman Institutional Review Board (Cincinnati, OH) approved the study protocol designed to compare treatment outcomes. Forty-three patients (22 experimental, 21 control) who met the inclusion criteria (ie, chronic, recalcitrant wounds) were randomized in a 1:1 ratio to the experimental or control group. The treatment regimen included a 12-week daily application of a wound gel (experimental [BlastX; Next Science]) or a broad-spectrum antimicrobial ointment (control [Neosporin + Pain Relief; Johnson & Johnson]). Patients ranged in age from 32 to 91 years (average age, 62 years). Wound age ranged from 1 month to 20 years (average wound age, 21.2 months). Wound size ranged from 1 cm² to 114 cm² (average wound area, 10 cm²). Patient comorbidities included diabetes mellitus (60%), peripheral arterial disease (40%), hypertension (35%), and obesity (26%). These comorbidities were not statistically significant factors affecting wound closure or healing rates when analyzed by ANOVA.

Study subjects were evaluated at time zero (baseline) as well as at 2, 4, 8, 12, and 16 weeks. Wounds were sampled at baseline and 4 weeks. Samples were submitted for DNA testing by proprietary genetic sequencing to identify microbial species and antibiotic resistance factors. All DNA testing was performed at the time of the original study; the current study is the first time reporting the results. DNA testing yielded bacterial load values of “No Results” (ie, clean), low, medium, and high. The clean results indicate that the bacteria were below the limit of detection of <10⁵ DNS copies/gram. Load values less than 10⁵ DNA copies/gram of bacteria in the sample were categorized as low, between 10⁵ and 10⁷ DNA copies/gram were categorized as medium, and greater than 10⁷ DNA copies/gram were categorized as high. All patients received the standard of care as a basic minimum and based on established wound etiology. At each visit, sharp wound debridement was performed, and wound measurements (area, volume, and depth) and images were obtained.

Statistics

Statistical analyses for this secondary submission were performed using Minitab (version 17.3.1; Minitab, Inc.) on the intent-to-treat population as in the original study. All patients who were enrolled and randomly allocated to treatment were included in the analysis and were analyzed in the groups to which they were randomized as well as in the crossover group where indicated. Statistical significance for quantitative independent variables was determined by ANOVA using a general linear model with factors as the treatment (experimental; control). Tukey pairwise comparisons were performed for grouping (P <.05). Treatment bars that do not fall under the same grouping bar are statistically distinct (P <.05; for the healed wounds and percent of wound closure). For categorical independent variables, a 2-sample proportion analysis was performed, with the P value determined by normal approximation.

Demographics

Kim et al⁶ analyzed the primary outcomes from the study (wound size reduction and wound closure), whereas the focus in the present study is on the bacterial bioburden testing results.

Biofilm analysis

Kim et al⁶ found that of the 90 bacterial and 4 fungal species analyzed, only 5% of samples tested positive for fungi. Seventeen bacteria were common to at least 10% of the patients. While early data from the original study found no statistically significant relationship between the bacterial species present with wound healing or healing rates, this secondary analysis found correlations between total bioburden levels and wound size reductions.

Wound size reduction and bacterial load

Wound size reduction was the primary endpoint in the original study. Kim et al⁶ found that in the experimental group the average decrease in wound area at 12 weeks was 72% ± 8 (standard deviation), which was statistically significant compared with control (P <.01). At the 4-week time point, the mean wound size in the experimental group was 43.6% smaller compared with baseline, while in the control group the mean wound size was only 29.9% smaller.

Data analysis indicated a correlation between lower bioburden at baseline and wound size reduction over time.

Interval plots of the bacterial load at baseline and 4 weeks and the percent wound size reduction at 4 weeks demonstrated more rapid healing in wounds with lower bacterial bioburden (Figures 1, 2). At baseline, average wound size reduction of 40% and 24%, respectively, was achieved in the low and medium bioburden groups, whereas in the high bioburden group wound size increased an average of 19%. Concerning bioburden at 4 weeks, 100% wound closure was achieved in the clean bioburden group, while the low bioburden group achieved an average wound reduction of 42% and the medium group experienced an average 4% increase in wound size.

These findings were statistically significant for the baseline (P =.087) and 4-week bacterial measures (P <.001). For the baseline data, grouping by the Tukey t test at a P value of less than .10 demonstrated that the low and high groups were statistically independent.

The low and medium and the medium and high groups were statistically equivalent. For the 4-week bacterial bioburden data, the low, medium, and high groups were all statistically independent according to the Tukey t test at a P value of less than .10.

An interval plot of the bacterial measure at baseline and 4 weeks compared with the percent wound size reduction at 12 weeks demonstrates that bioburden significantly affected wound closure (Figures 3, 4). Concerning the metric of bioburden at baseline, at 12 weeks the low and medium bioburden groups had achieved average wound size reduction of 53% and 36%, respectively; in contrast, wound size increased 19% in the high bioburden group (Figure 3). Concerning the metric of bioburden at 4 weeks, the clean group achieved 100% wound closure at 12 weeks, whereas the low and medium groups achieved an average wound size reduction of 58% and 6%, respectively (Figure 4).

These findings were statistically significant by ANOVA for the baseline (P =.09) and 4-week data (P =.004). Grouping by Tukey t test at a P value of less than .10 demonstrated that the low and high groups were statistically independent. The low and medium and the medium and high groups were statistically equivalent at this P value for the baseline bacterial bioburden data. The clean and low groups were statistically independent of the medium group for the 4-week bacterial bioburden data.

Healed wound percentages and bacterial load

To determine if the baseline bacterial load is indicative of wound healing, a correlation between the percentage of healed wounds and the bacterial load at 4 weeks and 12 weeks was calculated, as shown in Table 1. The bioburden at baseline and 4 weeks suggests that wounds with lower bioburden healed faster than those with higher bioburden. For the baseline bioburden data, wounds with low bioburden were more than twice as likely as those with high bioburden to be healed at 4 and 12 weeks. For the 4-week data, healing was achieved only in wounds with low bioburden.

For the percentage of wounds healed at 4 and 12 weeks, the results were not statistically significant for the bioburden at baseline values. However, they were statistically significant for the bioburden at 4 weeks values (P =.005 and P =.028 for the percentage of healed wounds at 4 weeks and 12 weeks, respectively).

Figure 5 and Figure 6 demonstrate this correlation between reduced bioburden and improved number of healed wounds.

Bacterial load group comparisons

The performance differences between treatments at baseline (time zero) and after 4 weeks of treatment are shown in Table 2. This table shows a trend towards reduced bacterial load over time with all treatments as well as superior results for the experimental group.

By combining the groups into high or low bioburden groups, as shown in Table 3, these differences can be examined using a 2-proportion analysis. This is supported because the clean and low bacterial samples were within the same group by the Tukey t test for wound closure and are supported by the grouping data showing healed versus unhealed wounds at 4 and 12 weeks. The low bioburden groups have bacterial colony counts less than 10⁵ CFU/g and are considered passing. Table 3 and Figure 7 demonstrate these differences.

In the control group, 1 out of 9 samples improved from a high bacterial load to a low bacterial load, compared with 5 out of 11 samples in the experimental group (Figure 7). The experimental group reduced the bioburden levels of 45% of the high bioburden group to the low bioburden group from baseline to 4 weeks, as opposed to the control group, which only reduced the bioburden levels of 11% of the samples from high to low bioburden.

An analysis comparing improvement in bacterial load between the control and experimental groups indicates a significant improvement in the experimental group (P =.06).

This study confirms that topical application of antimicrobial agents to chronic wounds in conjunction with debridement leads to reduced bacterial bioburden. Testing also demonstrates that superior wound size reduction and wound closure outcomes occur in patients with lower bioburden counts. These findings agree with other published studies. Versey et al¹⁴ recently reported that factors driving wound chronicity, including inflammation, are influenced by the wound microbiome. Rahim et al¹² found a correlation between bacterial load and wound healing.

The present analysis also demonstrates that using a biofilm-disrupting wound gel leads to superior reductions in bacterial load compared with traditional antimicrobials, with the product being 4 times more effective at reducing the bioburden from high to low levels. These results are consistent with the improved wound size reduction found in the original study⁶ and indicate that the superior bacterial load reduction by the wound gel product leads to improved wound size reduction over time.

A 2017 study by Johani et al¹⁵ demonstrated that 100% of nonhealing DFUs had polymicrobial biofilms and that wound infections begin as contaminated wounds with low bioburden levels that progress into multi-species sessile communities of microorganisms or biofilm communities of bacteria. In an analysis of a proprietary index developed and reported on by Fife et al,¹⁶ evidence of bioburden/infection and a worsening Wagner grade were reported to be significantly predictive variables in healing. Loesche et al¹⁷ noted that the “[m]icrobial burden of chronic wounds is believed to play an important role in impaired healing and the development of infection-related complications.” The relationship between wound size reduction and bacterial load is also discussed in a 2020 study by Cole and Coe.¹³ Using a point-of-care bacterial fluorescence signature (ie, bacteria do not provide a signature when the level of bacteria is ≤10⁴ CFU/g), those authors found that when a signature was no longer present, a switch to a healing trajectory in wounds was facilitated.

This analysis is limited by the retrospective use of bacterial bioburden testing results from the original study by Kim et al.⁶ The primary limitation of this study is that it cannot infer a direct causal link between products and changes in wound size reduction and healing. The correlation between bioburden levels and wound size reduction after treatment suggests a hypothesis, not proof.

Another limitation is that the patient populations between the experimental and control groups were not matched based on comorbidities. Statistically, the experimental and control groups were equivalent for the number of patients with diabetes and hypertension. There were higher percentages of patients in the experimental group with peripheral arterial disease and those classified as overweight than in the control group, which suggests that the experimental product was used in a group that was more challenging to treat.

This study is limited to using retrospective data analysis. Thus, the use of prospective design for future research must be considered. Further investigations are crucial to understanding the contribution of the microbiome to overall wound healing. Continuing to pose questions surrounding causality is essential to understanding how wound treatments may affect bioburden and facilitate improved patient outcomes and clinical practice.

The results of this study confirm that lower bacterial bioburden correlates with improved wound size reduction and closure. These results also demonstrate that use of a biofilm-disrupting wound gel combined with debridement is more effective than the control antibiotic ointment combined with debridement in reducing the bacterial bioburden of wounds. Wounds in the present study demonstrated measurable wound area reduction, increased wound closure rates, and a corresponding reduction in bioburden that reinforces the original research outcomes.

While many factors influence wound healing, the results obtained in this study may confirm that in chronic nonhealing wounds, bacterial loads affect wound size and healing. In this analysis, the decreasing bioburden seemed to correlate with wound size reduction and healing. There is increasing evidence that this decrease in wound size, as well as improved healing rates, confirm earlier findings that reducing bioburden and biofilm in wounds is an effective method to reduce wound size and improve healing.