The Association of Systemic Inflammation, Wound Bioburden and Total Bacterial Counts With Healing Outcomes in Older Adults With Chronic Venous Leg Ulcers
Funding: This study was supported by the National Institute of Nursing Research (R01:NR016986).
ABSTRACT
This substudy utilised data from a prospective, longitudinal study aimed to investigate associations between systemic inflammation, wound bioburden, total bacterial counts and wound healing outcomes over 8 weeks in older adults with chronic venous leg ulcers (CVLUs). Participants were receiving standardised weekly wound debridement. Blood and wound tissue samples were collected at baseline and Weeks 2, 4, 6 and 8, or until the wound was healed. Wound healing status was categorised by two parameters: healed versus nonhealed and healing versus nonhealing. A linear mixed model assessed associations among clinical and laboratory variables with wound healing status over time. Of the 117 participants, 47 (40%) had wounds that healed within the 8-week period. In nonhealed wounds, C-reactive protein (CRP) was positively associated with total bacterial counts (p < 0.001), wound bioburden (p = 0.01) and wound diameter (p = 0.004) over time. Total bacterial counts were positively associated with CRP (p = 0.023), interleukin-6 (IL-6) (p = 0.034) and tumour necrosis factor-alpha (TNF-α) (p < 0.001) in nonhealing wounds. The results suggest that CRP, IL-6 and TNF-α may be useful markers in predicting wound healing trajectories. Ongoing monitoring of inflammatory markers and bacterial counts could aid in assessing wound healing progress in older adults with CVLUs.
Summary
- Higher total bacterial counts in wounds at baseline can delay wound healing for patients undergoing weekly sharp debridement.
- CRP was positively associated with total bacterial counts (p < 0.001) and wound bioburden (p = 0.01), as well as wound diameter (p = 0.004) in nonhealed wounds over time.
- Monitoring the level of systemic CRP, TNF-α and IL-6 along with total bacterial counts can help predict CVLU wound healing trajectories in older adults.
- Positive associations were found between CRP, TNF-α, IL-6 and total bacterial counts over time in nonhealing wounds in CVLU among older adults.
1 Introduction
Chronic wounds are a significant cause of clinical, social and economic burdens on society and decreased quality of life, especially for the growing ageing population in the United States and globally [1-3]. Of chronic wounds, chronic venous leg ulcers (CVLUs) represent the most prevalent type of lower extremity wounds, accounting for 75%–80% of cases [4]. The prevalence of CVLUs increases with age [5, 6], with a frequency of 0.32% among individuals with a mean age of 39–79 [7], increasing to 3% in those over 80 years [6].
The latest report in the United States shows a 25.4% increase in both the number of individuals with CVLUs and the annual cost of treating them between 2014 and 2019. The cost has risen from around $600 million in 2014 to $1.13 billion in 2019 for CVLU treatment among Medicare beneficiaries [8]. The report also indicated that there is a significant increase in nonhealing CVLUs and that these wounds should be considered by clinicians as hard-to-heal wounds, due to poor healing rates and high reoccurrence. Although compression therapy and exercise are the most effective treatments up to now [5], there are no robust treatments for venous leg ulcers, which are difficult to heal. Indeed, it takes more than 4 weeks for wounds to reach a 40% area reduction, and 70% of chronic wounds require debridement [9]. The range of healing within 12 weeks was found to be 12.5%–88.3% based on 20 randomised controlled trials (RCTs) [10]. Additionally, approximately 20% of individuals experience a recurrence of their wounds within 3 months, over 50% within 12 months and 73% at 2 years [11].
Because of the poor rate of healing in venous leg ulcers, predicting wound healing trajectories is critical for making clinical decisions in wound treatment. The clinical and laboratory factors associated with CVLU wound healing are still not clear. The relationship of bacterial bioburden found in VLU has not been found to be directly related to healing outcomes but is associated with a longer duration of wound healing [12]. Studies indicate that it is necessary to understand microbial diversity and microbial load in wounds to determine the role of wound bioburden in healing outcomes [13]. Moreover, while diagnostic methods to determine wound infection have been developing [14], clinicians continue to rely on patients' symptom reports of wound infection for treatment [12].
It has been established that dysregulated inflammatory processes hinder wound healing. However, detailed mechanisms of immune cells at the wound site remain unclear [15]. Diverse cells such as macrophages, lymphocytes, mast cells, endothelial cells, astrocytes, fibroblasts and stromal cells release cytokines [16]. For instance, proinflammatory and anti-inflammatory cytokines released from the classically activated, called M1 (inflammatory macrophages and proinflammatory macrophages), and alternatively activated, called M2 (healing macrophages and anti-inflammatory macrophages) [17, 18] coordinate with other cells to promote wound healing processes such as cell proliferation, immune cell recruitment, angiogenesis and phagocytosis.
Increased proinflammatory cytokines and low anti-inflammatory cytokines are prominent in the microenvironment of chronic wounds [15, 19, 20]. Cytokines have been widely examined in the microenvironment of venous leg ulcers. Wound fluids are the most common specimen to examine cytokines, followed by wound tissue. Elevated levels of interleukin 1-alpha (IL-1α), interleukin 6 (IL-6), interleukin 8 (IL-8), tumour necrosis factor-alpha (TNF-α) and vascular endothelial growth factor (VEGF) in wound fluid have been reported in nonhealing venous leg ulcers [21].
Considering that wound healing is influenced by systemic factors such as age, comorbidities and nutritional status, along with local wound environments [22, 23], using the identified cytokines from localised wounds as biomarkers in clinical practice is challenging. Particularly, wound healing in individuals with CVLU is impacted by systemic inflammatory dysregulation [24], which is linked to multiple comorbidities (e.g., 54.8% of diabetes and 18.7% of hypertension among 1 225 278 Medicare enrolees who developed venous leg ulcers) [9].
C-reactive protein (CRP) includes proinflammatory and anti-inflammatory properties, rises in the acute phase of inflammation and fades when inflammation subsides. Because of these properties, levels of CRP are monitored in patients with a broad range of clinical conditions, including ischaemia, reperfusion injury, Alzheimer disease, age-related macular degeneration and immune thrombocytopenia [25]. Persistent elevation of CRP is observed in chronic infection [26]. Among patients with CVLU, systemic CRP is associated with symptoms such as poor sleep quality and mild fatigue [27], and wound-related symptoms such as pain, exudate and wound biofilm [28]. However, a link between systemic CRP and chronic wound healing trajectories has not been clearly identified in older adults. Therefore, we longitudinally observed inflammatory markers in blood along with wound bioburden and total bacterial counts in chronic wounds to describe the associations between systemic inflammation, wound bioburden and wound healing trajectories in older adults.
2 Material and Methods
2.1 Study Design
This study was a prospective longitudinal observational study.
2.2 Recruitment of Subjects
This study was approved by the Institutional Review Board of the University of *** (IRB201700566). Participants were enrolled after signing the written consent. Participants were recruited using convenience sampling from 2018 to 2022. All participants (N = 117) received weekly sharp debridement followed by standardised care at a university wound clinic. Compression dressings were applied using thin layered compression wraps unless contraindicated. To increase the study's feasibility by reducing dropout rate and based on the mean time wound healing rate (6.4 weeks) with a major healing time (4.3 weeks) [10], participants were followed up for 8 weeks. Considering that age 55 years or older is a risk factor for venous ulcers [29], the inclusion criteria were (1) subjects were aged 55 or older; (2) subjects had a venous leg ulcer confirmed by clinical diagnosis; (3) subjects had adequate arterial blood perfusion (ankle-brachial index between 0.7 and 1.3, inclusive or no occlusion as determined by Doppler measurement); (4) subjects had a chronic venous ulceration duration of more than 30 days [1, 30]; (5) subjects were cognitively intact as determined by a minimum score of 24 on the Mini-Mental State Examination. The exclusion criteria were (1) subjects who were undergoing kidney dialysis for renal failure; (2) subjects who were receiving immunosuppressant treatment, including systemic steroids or topical steroids in the wound within 4 weeks before the study; (3) subjects had a systemic infection; (4) subjects received chemotherapy within 4 weeks before study entry; (5) subjects had severe concomitant conditions, which required hospital admission; (6) subjects had immune suppression (HIV, transplant status) or autoimmune disorders. More detailed criteria can be found elsewhere [31, 32].
2.3 Data Collection
In the parent study, blood and wound tissue were collected at baseline and 2, 4, 6 and 8 weeks of treatment or until the wound was healed (if healed before 8 weeks, tissue and blood were not collected) between August 2018 and June 2022. This substudy utilises data from baseline, 4 and 8 weeks, at which times measurements of selected inflammatory markers were obtained. Each patient received follow-up care and wound debridement with the same physician during the treatment period.
2.4 Demographics
Information on age, gender, education, smoking habits and antibiotic use was collected from electronic health records and demographic questionnaires. The Charlson comorbidity index (CCI) [33] was calculated based on diagnostic information (ICD-10-CM). BMI was calculated by dividing a person's mass or weight (in kilogrammes) by height (metres squared) to determine underweight, normal weight and obesity.
2.5 Wound Diameter
Wound size, the wound perimeter, area and volume were measured by the Silhouette (ARANZ Medical) and validated by a certified wound nurse expert, and wound diameter was calculated by conversion of wound area measurements to the equivalent circular diameter. The Silhouette is a hand-held device that minimises variability between clinicians who measure wounds by providing laser-guided accuracy to the level of millimetres of length, width and depth of the wound, with software providing other calculations, including volume. The Silhouette has demonstrated a high level of reliability, validity and reproducibility across studies [34].
2.6 Determination of Wound Healing Status
Two parameters were used to denote wound healing status. Alternate endpoints have been suggested for clinical trials of wound healing. The first measure, ‘healed’ versus ‘non-healed’ used a physician rating of healing. Healed wounds refer to wounds that were identified as clinically healed by the physician's clinical judgement at some point during the 8-week study period. Nonhealed wounds were not declared as clinically healed during the 8 weeks. Forty-seven participants had healed wounds during the eight-week study period and 70 participants had nonhealed wounds at 8 weeks.
The second measure, ‘healing’ versus ‘non-healing’, used a formula that calculated linear healing per unit of time, using the approach described by Gorin et al. (1996). There are two benefits of this approach: first, monitoring the healing rate over several weeks offers a more rapid indicator, whereas the time to complete closure can take months. Secondly, it enables a self-controlled model, where postintervention data from each participant can be paired with their individual preintervention data, creating a reliable control group within each subject [35]. Healing wounds were determined by classifying as having a linear healing slope of < 0 and nonhealing wounds were classified by a linear healing slope of 0 or above. During the 8 weeks of the study period, there were 96 healing wounds and 21 nonhealing wounds.
2.7 Wound Tissue Specimen Collection
The physician performed sharp debridement followed by clinical judgment to accomplish wound healing, and a nurse researcher collected the entire debrided wound tissue during the procedure. A dermal curette was primarily used, with scissors, forceps or a scalpel employed as needed. Debridement of the edges of wounds was performed first, followed by the base of the wound. If a patient had multiple wounds, tissue was obtained from the largest wound. The debrided material was subdivided, with a portion reserved for microscopy and another portion for bacterial enumeration.
2.8 Wound Bioburden
Bioburden is a general term covering both bacterial load and biofilm bacteria from wounds, which is not obtained from culturing wound tissue. Sub-samples of wound tissue debridement material were fixed with 4% paraformaldehyde in phosphate-buffered saline, cryo-embedded in optimum cutting temperature compound (OCT, Tissue-Tek, Fisher Healthcare, USA) and stored at −80°C. Thin sections, 5 μm thick, were cut at −20°C using a Leica CM 1850 Cryostat (Leica, Wetzlar, Germany) and placed on plus slides. The specimens were stained using Sytox green and wheat germ agglutinin conjugated with Texas Red (WGR-TR) (Life Technologies, Carlsbad, CA), and examined using a Leica SP5 confocal scanning laser microscope. Specimens were scored for bioburden using a 6-point scale; 0 = no bacteria observed, 1 = single individual microorganisms, 2 = small microcolonies (10–100 cells), 3 = large microcolonies (> 100 cells), 4 = continuous biofilm and 5 = thick (> 10 um) continuous biofilm [36].
2.9 Total Bacterial Counts
The level of total viable bacteria, referred to as total bacterial counts in this study, was measured using minor modifications of the basic technique described by Phillips, Yang and Schultz [37]. Briefly, a sub-sample of the curettage specimen collected from patients was immediately placed in a sterile 15 mL tube with 1.5 mL PBS containing 5 ppm Tween 20 and transported to the laboratory at 4°C using cold packs. Samples were vortexed for 1 min at high speed, which dislodged both planktonic and biofilm bacteria from the curettage tissue sample. The level of total bacteria was measured in an aliquot (400 μL) of the vortexed sample by plating following serial dilution onto TSA agar plates using a spiral plating method of nondiluted, 102-fold diluted and 105-fold diluted samples (EasySpiral Dilute, Interscience, Saint Nom, France). After incubation for 24 h at 32°C, the number of colonies was counted using an EasyScan 300 (Interscience, Saint Nom, France) in the dilution zones that contained between 15 and 300 CFU, or 0 CFU if no colonies were present on the nondiluted plates. Values of total viable bacteria were expressed as log10(CFU). Total bacterial counts were averaged across sample time points to calculate a single total bacterial number for each subject.
2.10 Blood Specimen Collection
At baseline and biweekly, participants had 3 mL of blood samples collected by venipuncture at the clinic for systemic inflammatory measures. Following the collection in a lavender-topped EDTA tube, blood specimens (3 mL) were placed in a biohazard container and transported to the university laboratory, where they were processed (within 4 h) [38] and stored at −80°C in the biorepository in preparation for batch processing for the measure of CRP and cytokine concentrations.
2.11 Systemic Cytokines
Systemic cytokines TNF-α, IL-1β, IL-4, IL-6 and INF-γ were analysed using a Magnetic Luminex Performance Assay (Human HS Cytokine Premixed Kit A, cat # FCSTM09-05; R&D Systems, Minneapolis, MN) according to the manufacturer's instructions using a Millipore MILLIPLEX Analyser, Luminex xMAP 100/200 Technology, with the Luminex xPONENT versus 4.3.1 data acquisition software (MilliporeSigma, Burlington, MA). Normal IL-6 in the healthy blood ranged from 0 to 43.5 pg/mL [39].
2.12 Systemic CRP
CRP is a clinically validated marker indicating inflammation and it is useful in determining the degree of systemic inflammation, progression or the effectiveness of treatments. Normal CRP concentration in healthy human blood is usually lower than 10 mg/L, slightly increasing with ageing. Higher levels are found in mild inflammation and viral infections (10–40 mg/L), active inflammation, bacterial infection (40–200 mg/L) and severe bacterial infections. Here, we measured systemic CRP using ELISA, a commercial assay, according to manufacturer's instructions (DuoSet ELISA Development System, Human CRP, R&D Systems, Minneapolis, MN).
2.13 Statistical Analysis
The analysis included data collected from 117 participants at baseline, Week 4 and Week 8. Based on 47 healed and 70 not healed, there was 80% power to detect a medium effect size (Cohen's d) of 0.55 using a Wilcoxon nonparametric test for difference in means [40]. Distributions for each variable were examined using descriptive statistics appropriate for measurement level. Due to the skewed distributions of the wound area and total bacteria, bioburden, CRP and cytokines (TNF-α, IL-1β, lL-4, IL-6 and lFN-γ) variables, a log base 10 transform was applied to normalise the distribution of values. Patterns of missingness were evaluated using the nonparametric omnibus test for missing completely at random (MCAR) as implemented by TestMCARNormality within the MissMech R package [41-44]. Due to censoring (cytokines were present at very low levels or not detected, and these levels were below the limit of detection for the methods used), IL1b (> 50%), IL4 (> 50%) and IFN-γ (100%) values were excluded from the analyses. Separate generalised linear mixed models (GLMM) were used for each of the outcomes (total bacterial counts, bioburden and wound diameter). As the GLMM approach used all available data for each participant at each time point, no data imputation was utilised; 90% of data for testing CRP association with wound diameter were nonmissing, as was 65% of data for testing CRP with bacterial load. The lower nonmissing yield for bacterial load was primarily a result of insufficient wound material to perform bioburden measurements. Analyses were conducted using SAS version 9.4 and R version 4.3. Since IL-1β, IL-4 and IFN-γ were not included in the analyses, we did not present their values in the tables.
3 Results
3.1 Participants Characteristics
A total of 117 participants were enrolled. 47 (40%) were healed, while 70 (60%) were not healed in 8 weeks. Table 1 indicates sample characteristics of the healed and nonhealed groups. Among participants' characteristics, the mean of wound duration differed at baseline (161.7 [SD = 256.3] vs. 428.5 [SD = 587.9], p = 0.004) between the healed and nonhealed group.
| Mean (SD) or frequency (%) | p | Range | ||
|---|---|---|---|---|
| Healed 47 (40%) | Nonhealed 70 (60%) | |||
| Age | 73.8 (10.0) | 70.3 (9.1) | 0.059a | 55–92 |
| Wound duration (days) | 161.7 (256.3) | 428.5 (587.9) | 0.004b | 20–2944 |
| BMI | 32.3 (7.3) | 35.3 (14.1) | 0.619b | 15.4–84 |
| Charlson index | 5.6 (1.8) | 5.9 (2.0) | 0.620b | 2–11 |
| Antibiotic use | 0.015c | NA | ||
|
Yes No |
22 (47%) 25 (53%) |
49 (70%) 21 (30%) |
||
| Race | 0.371c | NA | ||
|
White African-American |
40 (85%) 7 (15%) |
54 (77%) 16 (23%) |
||
| Sex | 0.180c | NA | ||
|
Male Female |
22 (47%) 25 (53%) |
42 (60%) 20 (40%) |
||
| Smoking | 0.120c | NA | ||
|
Never smoked Former smoker Current smoker |
26 (55%) 20 (43%) 1 (2%) |
32 (46%) 29 (41%) 9 (13%) |
||
| Education | 0.737c | NA | ||
|
Elementary Some HS GED/HS grad Some college College graduate Prefer no answer |
0 (0%) 3 (6%) 12 (25)% 17 (36%) 14 (30%) 1 (2%) |
3 (4%) 7 (10%) 17 (24%) 20 (29%) 22 (31%) 1 (1%) |
||
| Income | 0.414c | NA | ||
|
< $25 K $25 K–$34 K $35 K–$44 K $45 K–$54 K $55 K–$64 K $65 K–$74 K $75 K–$84 K $85 K–$94 K > $95 K Prefer no answer |
9 (19%) 7 (15%) 3 (6%) 1 (2%) 3 (6%) 3 (6%) 2 (4%) 3 (6%) 5 (11%) 11 (23%) |
22 (31%) 5 (7%) 2 (3%) 5 (7%) 2 (3%) 3 (4%) 2 (3%) 2 (3%) 4 (6%) 23 (33%) |
||
- a Based on Welch T-test.
- b Based on Wilcoxon rank-sum test.
- c Based on Monte Carlo simulation with 2000 replications.
3.2 Differences in Wound Factors (Total Bacterial Counts, Bioburden and Wound Diameter) and Systemic Inflammation Between the Healed and Nonhealed CVLU Groups
Systemic CRP(p = 0.049), IL-6 (p = 0.034), total bacterial counts (p = 0.033) and wound diameter (p < 0.001) differed at baseline between groups, with the nonhealed participants having larger values of CRP, IL-6, total bacterial counts and wound diameter. Wound diameter decreased during the study period in both groups. The nonhealed group had relatively larger wounds than the healed group during the entire study period. CRP differed significantly from baseline to Week 4 between the healed and the nonhealed group, which increased in the nonhealed group. However, a difference in CRP between groups was not significant at Week 8 due to decreasing values in the nonhealed group. There were no differences in the levels of TNF-α (p = 0.181–0.591) and bioburden (p = 0.372–0.736) between healed and nonhealed groups during the 8 weeks of the study period. See Table 2 for details.
| Healed | Nonhealed | p a | |||
|---|---|---|---|---|---|
| Biofilm at baseline | 0.348 | ||||
| No | |||||
| 14 (52%) | 21 (38%) | ||||
| Yes | |||||
| 13 (48%) | 34 (62%) | ||||
| Mean (SD) | Range | Mean (SD) | Range | p b | |
| LOG10 (CRP) | |||||
| Baseline | 0.59 (0.49) | −0.70–2.2 | 0.76 (0.52) | −0.42–2.13 | 0.049 |
| Week 4 | 0.56 (0.41) | −0.31–1.42 | 0.81 (0.48) | −0.38–1.94 | 0.010 |
| Week 8 | 0.58 (0.39) | −0.32–1.04 | 0.67 (0.54) | −0.46–2.16 | 0.771 |
| Log10 (TNF-α) | |||||
| Baseline | 1.0 (0.21) | 0.46–1.51 | 0.99 (0.18) | 0.58–1.39 | 0.591 |
| Week 4 | 0.99 (0.21) | 0.40–1.32 | 1.00 (0.18) | 0.57–1.43 | 0.804 |
| Week 8 | 0.89 (0.24) | 0.42–1.35 | 0.98 (0.17) | 0.62–1.33 | 0.181 |
| Log10 (IL-6) | |||||
| Baseline | 0.20 (0.26) | −0.36–0.70 | 0.36 (0.32) | −0.23–1.23 | 0.034 |
| Week 4 | 0.22 (0.23) | −0.12–0.75 | 0.33 (0.28) | −0.19–0.91 | 0.128 |
| Week 8 | 0.24 (0.47) | −0.14–1.22 | 0.35 (0.35) | −0.23–1.48 | 0.161 |
| LOG10 (total bacterial counts) | |||||
| Baseline | 3.2 (2.1) | 0–7.2 | 4.2 (2.0) | 0–7.7 | 0.033 |
| Week 4 | 3.3 (2.3) | 0–6.6 | 4.3 (2.3) | 0–8.1 | 0.086 |
| Week 8 | NA 0 (0) | NA 0–0 | 3.8 (2.2) | 0–7.7 | NA |
| LOG10 (bioburden) | |||||
| Baseline | 1.6 (2.0) | 0–6.0 | 1.8 (2.1) | 0–6.8 | 0.736 |
| Week 4 | 1.6 (2.0) | 0–6.3 | 2.2 (2.3) | 0–7.4 | 0.372 |
| Week 8 |
NA NA |
2.0 (2.3) | 0–7.5 |
NA NA |
|
| Wound diameter | |||||
| Baseline | 24.2 (24.4) | 2.0–166.4 | 48.0 (42.8) | 5.5–249.6 | < 0.001 |
| Week 4 | 14.4 (25.1) | 0–154.5 | 43.1 (37.2) | 7.3–212.2 | < 0.001 |
| Week 8 | NA 0 (0) | NA 0–0 | 36.3 (36.0) | 2.8–153.1 | NA |
- Note: CRP (mg/L); TNF-α (pg/mL); IL-6 (pg/mL); wound diameter (mm).
- a Based on Monte Carlo simulation with 2000 replications.
- b Based on Wilcoxon rank-sum test.
3.3 Association Among Wound Factors (Total Bacterial Counts, Bioburden and Wound Diameter), Systemic CRP and Cytokines in Healed Wounds
Total bacteria, bioburden and wound diameter did not have an association with systemic inflammatory markers such as CRP, TNF-α and IL-6 (0.28 < p < 0.93) in the regression model. In the full-mixed model, CRP (p = 0.001) was negatively associated with wound diameter while TNF-α (p = 0.021) and IL-6 (p < 0.001) were positively associated with wound diameter. See Table 3 for details.
| Dependent | Independent | b a | p | R 2 marginal b | R 2 conditional c |
|---|---|---|---|---|---|
| Log10 total bacterial counts | Log10 CRP | −0.285 | 0.68 | 0.003 | 0.494 |
| Log10 total bacterial counts | Log10 TNF-a | 1.44 | 0.49 | 0.012 | 0.502 |
| Log10 total bacterial counts | Log10 IL-6 | −1.23 | 0.33 | 0.021 | 0.551 |
| Log10 bioburden | Log10 CRP | −0.691 | 0.28 | 0.024 | 0.379 |
| Log10 bioburden | Log10 TNF-a | 0.224 | 0.91 | < 0.001 | 0.342 |
| Log10 bioburden | Log10 IL-6 | −0.917 | 0.44 | 0.014 | 0.370 |
| Log10 wound diameter | Log10 CRP | −0.045 | 0.74 | 0.001 | 0.001 |
| Log10 wound diameter | Log10 TNF-a | 0.399 | 0.25 | 0.021 | 0.021 |
| Log10 wound diameter | Log10 IL-6 | 0.022 | 0.93 | < 0.001 | < 0.001 |
- Note: The healed wound total did not count any loss due to missing other variables in models.
- a Estimated regression weight.
- b R2 is explained by fixed effects.
- c R2 is explained by full-mixed model.
3.4 Association Among Wound Factors (Total Bacterial Counts, Bioburden and Wound Diameter), Systemic CRP and Cytokines in Nonhealed Wounds
CRP was significantly associated with total bacteria (p < 0.001), bioburden (p < 0.01) and wound diameter (p < 0.004) from the regression model. Higher CRP is associated with higher number of total bacteria, higher number of bioburden and larger wound diameter in nonhealed wounds. None of the cytokines were associated with wound factors. See Table 4 for details.
| Dependent | Independent | b a | p | R 2 marginal b | R 2 conditional c |
|---|---|---|---|---|---|
| Log10 total bacterial counts | Log10 CRP | 1.42 | < 0.001 | 0.125 | 0.446 |
| Log10 total bacterial counts | Log10 TNF-a | −0.270 | 0.84 | < 0.001 | 0.451 |
| Log10 total bacterial counts | Log10 IL-6 | 0.982 | 0.17 | 0.021 | 0.435 |
| Log10 bioburden | Log10 CRP | 0.994 | 0.01 | 0.057 | 0.478 |
| Log10 bioburden | Log10 TNF-a | 0.932 | 0.53 | 0.005 | 0.501 |
| Log10 bioburden | Log10 IL-6 | 0.956 | 0.22 | 0.015 | 0.509 |
| Log10 wound diameter | Log10 CRP | 0.134 | 0.004 | 0.040 | 0.758 |
| Log10 wound diameter | Log10 TNF-a | 0.165 | 0.411 | 0.006 | 0.761 |
| Log10 wound diameter | Log10 IL-6 | 0.080 | 0.37 | 0.005 | 0.760 |
- a Estimated regression weight.
- b R2 is explained by fixed effects.
- c R2 is explained by full-mixed model.
3.5 Association Among Wound Factors (Total Bacterial Counts, Bioburden and Wound Diameter), Systemic CRP and Cytokines in Healing Wounds
Among healing wounds (n = 96), only CRP was associated with total bacteria (p < 0.004) and wound diameter (p < 0.004) from the regression model. A higher number of total bacteria and larger wound diameter were associated with higher systemic CRP in healing wounds. In healing wounds, there is no association with CRP and bioburden. See Table 5 for details. None of the cytokines were associated with wound factors.
| Dependent | Independent | b a | p | R 2 marginal b | R 2 conditional c |
|---|---|---|---|---|---|
| Log10 total bacterial counts | Log10 CRP | 1.12 | 0.004 | 0.064 | 0.491 |
| Log10 total bacterial counts | Log10 TNF-a | −0.152 | 0.91 | < 0.001 | 0.553 |
| Log10 total bacterial counts | Log10 IL-6 | 0.364 | 0.59 | 0.002 | 0.542 |
| Log10 bioburden | Log10 CRP | 0.518 | 0.18 | 0.014 | 0.427 |
| Log10 bioburden | Log10 TNF-a | 0.759 | 0.58 | 0.003 | 0.482 |
| Log10 bioburden | Log10 IL-6 | 0.561 | 0.43 | 0.005 | 0.478 |
| Log10 wound diameter | Log10 CRP | 0.209 | 0.006 | 0.038 | 0.274 |
| Log10 wound diameter | Log10 TNF-a | 0.408 | 0.13 | 0.020 | 0.249 |
| Log10 wound diameter | Log10 IL-6 | 0.260 | 0.10 | 0.021 | 0.219 |
- Note: Healing wounds were accounted for (slope < 0).
- a Estimated regression weight.
- b R2 is explained by fixed effects.
- c R2 is explained by full-mixed model.
3.6 Association Among Wound Factors (Total Bacterial Counts, Bioburden and Wound Diameter), Systemic CRP and Cytokines in Nonhealing Wounds
Among 21 nonhealing wounds during the 8 weeks of the study period, none of the CRP or cytokines was associated with any of the wound factors from the regression model in nonhealing wounds. However, in the full-mixed model, CRP (p = 0.023), TNF-α (p < 0.001) and IL-6 (p = 0.034) were positively associated with total bacteria. See Table 6 for details.
| Dependent | Independent | b a | p | R 2 marginal b | R 2 conditional c |
|---|---|---|---|---|---|
| Log10 total bacterial counts | Log10 CRP | 0.490 | 0.36 | 0.023 | 0.023 |
| Log10 total bacterial counts | log10 TNF-a | 0.288 | 0.88 | < 0.001 | < 0.001 |
| Log10 total bacterial counts | Log10 IL-6 | 1.13 | 0.34 | 0.034 | 0.034 |
| Log10 bioburden | Log10 CRP | 1.03 | 0.12 | 0.065 | 0.515 |
| Log10 bioburden | Log10 TNF-a | −0.052 | 0.98 | < 0.001 | 0.449 |
| Log10 bioburden | Log10 IL-6 | 0.693 | 0.66 | 0.009 | 0.443 |
| Log10 wound diameter | Log10 CRP | 0.044 | 0.41 | 0.005 | 0.878 |
| Log10 wound diameter | Log10 TNF-a | −0.072 | 0.72 | 0.001 | 0.913 |
| Log10 wound diameter | Log10 IL-6 | −0.083 | 0.42 | 0.005 | 0.919 |
- Note: Nonhealing was accounted for (slope 0 or greater).
- a Estimated regression weight.
- b R2 is explained by fixed effects.
- c R2 is explained by full-mixed model.
4 Discussion
The study presented here describes the link between systemic inflammation, including CRP, TNF-α, IL-6, total bacterial counts and wound bioburden, in the trajectories of CVLU healing in older adults. Our study reveals that CRP is positively associated with wound diameter, bioburden and total bacterial counts over time in nonhealing wounds. Given that CRP is a frequently utilised measure in clinical settings for different inflammatory conditions, the consideration for adding this well-validated and reliable measure to wound care may be indicated.
4.1 The Association Between Systemic Inflammation and Wound Healing in Healed Wounds and Healing Wounds
Although we observed the differences in systemic levels of IL-6, and CRP between healed and nonhealed wounds (Table 2), only the CRP was associated with total bacterial counts and wound diameter in healing wounds from the regression model (Table 5). The previous studies examined increased systemic IL-1α, IL-6, IL-8, TNF-α and VEGF with nonhealing venous leg ulcers. However, none of these systemic inflammatory markers were significantly associated with healing venous leg ulcers [24, 45-47]. In nonhealing wounds, excessive proinflammatory cytokines are found [20], but systemic proinflammatory cytokines have not been identified in wound healing. In comparison, higher serum CRP has been reported in chronic surgical wounds [48, 49] and decreased serum CRP in healing diabetic wounds [50].
According to the study results, the regression model shows that the level of systemic CRP increases in line with the total number of bacteria and the size of the wound in healing wounds (Table 5). Additionally, the systemic CRP, TNF-α and IL-6 are associated with wound diameter in healed wounds, as indicated by the full-mixed model (Table 3). To the best of our knowledge, only few studies have longitudinally observed the characteristics of systemic inflammation and chronic wound healing. Notably, there have been no reports discussing a significant association between systemic CRP and wound healing in venous leg ulcers. Systemic CRP responds to the total bacterial counts and the wound diameter of healing wounds in CVLUs. Based on the results, longitudinal examinations of wound bioburden and blood CRP may help to predict wound healing trajectories during wound treatment.
4.2 The Association Between Systemic Inflammation and Nonhealed Wound and Nonhealing Wounds
Systemic CRP was associated with total bacterial counts, bioburden and wound diameter in nonhealed wounds, as shown in the regression model (Table 4). However, there was no significant association between CRP and wound factors from the fixed or full-mixed model. However, the CRP value decreased by Week 8 (Table 2), which probably indicated disruption of the normal healing process regulated by adequate inflammatory responses to multiple wound factors. It is consistent that systemic CRP inversely affects wound healing [51]. None of the systemic inflammatory markers were associated with any wound factors in nonhealing wounds from the regression model (Table 6). In the full-mixed model, proinflammatory cytokines TNF-α and IL-6 were significantly associated with total bacterial counts (Table 6).
An increase in TNF-α in the blood was observed in nonhealing venous leg ulcers [45], while there is no report of the association between systemic IL-6 and nonhealing venous leg ulcers. An increased IL-6 in wound fluid was examined in both healed and nonhealed wounds of venous leg ulcers [52, 53]. However, we discovered that elevated IL-6 levels in the systemic were linked with nonhealing wounds, which is the first instance of this association being described in venous leg ulcers. Considering the results collectively from nonhealed and nonhealing wounds from this study, total bacterial counts play a significant role in healing trajectories that increased proinflammatory cytokines which contributes to wounds being chronic and eventually holt normal wound healing process.
The meta-analysis indicates that the relationship between wound bioburden and wound healing is inconclusive [54]. Furthermore, the relationship between wound factors such as infection status, bacterial diversity and specific colonies that impact wound healing and healing outcomes in venous leg ulcers remains unclear. This study elucidates the association between wound healing outcomes and bioburden in CVLUs. Further study is needed to explore the effect of the diversity of wound bacteria on wound healing outcomes or the association between them.
Further research is needed to describe the association between wound inflammatory markers, systemic inflammatory markers and wound healing trajectory.
Local and systemic inflammation is associated with behaviours such as fatigue [55]. In addition, local bioburden is positively associated with fatigue and pain, and negatively associated with healed wounds [28]. Erythrocyte sedimentation rate (ESR) and CRP in blood serve as markers for wound infection [56]. However, as far as we know, the association between local and systemic inflammation has never been described in the wound population.
Furthermore, there is a lack of research on how their associations affect the wound-healing trajectory. Hence, we suggest more research to explore the link between local wound bioburden, local wound inflammation, systemic inflammation, individual behaviours associated with wound healing and wound healing trajectory. Further research investigating other cytokines and ESR in wound healing trajectories is beneficial for delineating associations or influences between local, systemic inflammation and wound healing. The results may provide a clue that alternate treatment options can break a vicious cycle of wounds leading to chronic wounds. Systemic inflammatory markers, such as CRP, can be valuable for indirectly measuring wound bioburden during the initial assessment, which helps guide wound treatment strategies.
4.3 The Limitations of the Study
Due to the small sample size, IL-1β, IL-4 and IFN-γ were censored, which we did not include in the discussion. The study's results may differ if enrolment expanded to additional clinics with different treatment protocols. The study's findings may have limited generalisability because the participants were recruited from a regional clinic and followed the standard treatment protocol of the clinic. The participants were only observed for 8 weeks during the study period, which could lead to different healing outcomes compared to other lengths of observation periods. We did not consider the potential impact of treatment received before the study's enrolment on wound healing outcomes. Lastly, different measurement techniques of CRP and cytokines may impact the study results.
5 Conclusion
The results highlight the importance of maintaining adequate local and systemic inflammatory responses during the healing process by eliminating wound bioburden, which is the most modifiable factor promoting wound healing. Further research is needed to validate the association between wound bioburdens, local and systemic inflammatory responses and wound healing trajectories with a large sample from diverse settings. This validation will enhance treatment strategies for CVLUs.
Acknowledgements
J.K.'s effort for this work was supported by the National Research Foundation of Korea, Grant No. NRF-2022R1C1C1006659.
Ethics Statement
This study was approved by the Institutional Review Board of the University of Florida (UF). (IRB201700566).
Consent
All participants were enrolled after they signed a written informed consent form.
Conflicts of Interest
The authors declare no conflicts of interest.
Open Research
Data Availability Statement
The authors have nothing to report.
