Composition of Wound Fluid from Pressure Ulcers Treated with Negative Pressure Wound Therapy Using V.A.C.® Therapy in Home Healt

Pressure ulcers develop when persistent pressure on soft tissue over a bony site obstructs capillary flow, leading to tissue necrosis.¹ Physiological response to tissue injury initiates a cascade of overlapping events that may last from days to months, depending on the ability of the wound to proceed through the healing process in an orderly and timely manner. An acute wound becomes chronic when it fails to heal properly in an appropriate timeframe. Some of the key regulators of wound healing include pro-inflammatory cytokines, growth factors, proteases, and protease inhibitors.^2,3 Wound chronicity is thought to be, in part, a result of disruption in the temporal sequence and the concentrations of these biomolecules.
Negative pressure wound therapy (NPWT; V.A.C.® Therapy, KCI, San Antonio, Tex) has been shown to help promote healing of wounds, including pressure ulcers.^4–9 One postulate is that modulation of key mediators of wound healing contribute to the clinical success associated with NPWT. Previous studies in pigs have shown that NPWT can have an effect on systemic levels of inflammatory cytokines within the first few hours of application.^10,11 Stechmiller et al¹² showed that wound fluid from NPWT-treated pressure ulcers in patients in an acute care setting exhibited significant decrease in the levels of tumor necrosis factor (TNF)-α from pre-therapy levels but presented no differences in protease levels. It is thought that the inpatients in that study had acute inflammation (as demonstrated by elevated initial TNF-α levels) but may have had lower initial levels of proteases because of better attention to debridement, treatment compliance, and nutritional support than in non-inpatient care settings. Thus, the current study focused on pressure ulcer patients in home care and extended care settings. The purpose of this pilot study was to test the hypothesis that pro-inflammatory cytokine, matrix metalloproteinase (MMPs), and tissue inhibitor of metalloproteinase (TIMP) levels in wound fluid from pressure ulcers treated with NPWT change over a period of 7 days.

Materials and Methods

Eight adults with open and draining Stage III or IV pressure ulcers were recruited after obtaining informed consent. Details of the inclusion and exclusion criteria are provided in Table 1.
A prospective, repeated measures design was utilized to investigate the concentrations of TNF α, interleukin (IL)-1b, MMP-3, MMP-9, and TIMP-1 in wound fluids of patients with pressure ulcers on NPWT. Wound fluid was collected at baseline (Day 0; prior to initiation of NPWT) and at Days 1, 3, and 7 of NPWT. Thus, each subject served as his/her own control.¹³
Prior to initiation of NPWT (-125 mmHg, continuous therapy; V.A.C.® Classic or Mini V.A.C.® Therapy units, GranuFoam® Dressings, KCI, San Antonio, Tex), wound fluid was collected from under a polyurethane occlusive dressing (Tegaderm™, 3M, St. Paul, Minn) after allowing fluid to collect in the wound for 1 to 2 hours, per a previously described procedure.¹⁴ Wound fluid was collected directly from the NPWT system tubing during the remainder of the study. These specimens were immediately centrifuged (Eppendorf Centrifuge, Westbury, NY; 6800 g, 15 minutes), and the supernatants were subdivided into separate microvials and stored at -80°C until analyzed. If the day of fluid collection coincided with the day of dressing/tubing changes, wound fluid was collected from the NPWT system at least 2 hours after dressing and/or tubing changes to minimize potential contamination from blood during dressing changes. In a previous study, exposure of wound fluid to the GranuFoam Dressing, V.A.C. tubing, or incubation at room temperature for 6 hours did not affect concentrations of the analytes evaluated in the study.¹⁵
Analysis of cytokines and proteases in wound fluid. Quantitative measurements were performed in triplicate using commercially available (Amersham Biosciences, Buckinghamshire, UK) enzyme-linked immunosorbent assay (ELISA) kits for TNF α (TNF α Human Biotrak Assay, RPN2758), IL-1β (IL-1β Human Biotrak Assay, RPN2751), MMP-3 (MMP-3 Human Biotrak Assay ELISA, RPN 2613; measures pro-MMP-3, active MMP-3, and complexed MMP-3), MMP-9 (MMP-9 Biotrak Activity Assay, RPN 2634; measures total levels of MMP-9), and TIMP-1 (TIMP-1 Human Biotrak Assay, RPN 2611).
Statistical analyses. Data were tested for suitability of parametric analysis with the Kolmogorov-Smirnov test for normal distribution and the Levine Median test for equal variance. Depending on the outcome of these tests, either parametric or non-parametric 1-way repeated measure analyses of variance (ANOVA) were used to determine differences among the various time points. If differences were detected, multiple comparisons were done with the student-Newman-Keuls test; P < 0.05 was considered statistically significant.

Results

Twenty subjects were considered for participation in this study; 11 subjects did not meet the study criteria. One subject was dropped from the study because of noncompliance with the protocol. Thus, a total of 8 subjects were enrolled. Baseline subject characteristics were age, gender, diabetes status, and stage of pressure ulcer (Table 2).
Levels of MMP-3, MMP-9, and MMP-3:TIMP-1 ratio had a statistically significant decrease from Day 0 to Day 1. This decrease was sustained at Day 3 for all of these 3 metrics and at Day 7 for MMP-3 and MMP-3:TIMP-1 ratios (Figures 1–3, Table 3). There were no statistically significant differences among Days 1, 3, and 7. There were no statistically significant differences detected for TIMP-1, MMP-9:TIMP-1, IL-1β, or TNF α levels at any of the time points when compared to baseline or among the time points (Figures 4–7, Table 3).

Discussion

While NPWT has been shown to be effective in promoting wound healing,^4–9 the physiological mechanisms associated with this therapy are continuing to be explored. One possible mode of action is a general intervention in the inflammatory process. In this pilot study, changes in concentrations of inflammatory cytokines and proteases were proposed to occur over a period of 7 days in wound fluids from pressure ulcer patients during NPWT. A synopsis of the significance of the bioanalytes examined in this study is presented in Table 4.
While inflammation is an important and necessary part of the wound healing process, in wounds, such as pressure ulcers, there is a persistent inflammatory stimulus that leads to chronicity. Chronic inflammation is associated with a vicious cycle where inflammatory cells secrete cytokines that in turn attract more inflammatory cells.^16,17 At the molecular level, this has been shown to be associated with an imbalance between the concentration of proteases (eg, MMP-3, MMP-9), protease inhibitors (eg, TIMP-1), and pro-inflammatory cytokines (eg, TNF α, IL-1β).^2,18,19 Fluid analyses from healing wounds have shown that healing wounds have high levels of mitogenic activity, low concentrations of inflammatory cytokines, low levels of proteases, and high levels of growth factors.^16,19,20 In contrast, chronic wounds are characterized by low mitotic activity, high levels of inflammatory cytokines, high levels of proteases, and low levels of growth factors.
Matrix metalloproteinases are present in all phases of wound healing and promote the repair process by degrading extracellular matrix, inducing cell migration, and stimulating the remodeling or repair process. MMP-3 expression has been shown in macrophages as well as specific subsets of keratinocytes in the proliferating cell sub-population in normal and chronic wounds.^21,22 MMP-3 (also known as stromelysin-1) degrades a number of substrates, including type IV collagen, fibronectin, gelatin, and laminin, and is important in activating other pro-MMPs.^23,24 MMP-3 knockout mice show delayed wound healing, resulting from impaired early wound contraction.²⁵ Similarly, MMP-9 (also known as gelatinase-B) knockout mice have been shown to have defective remodeling of extracellular matrix (particularly fibrin) along with increased rates of re-epithelization.²⁶ On the other hand, prolonged and elevated levels of MMPs are associated with chronic wounds.^{2,16,18,19,20,27} Matrix metalloproteinases continually break down the extracellular matrix components, growth factors, and growth factor receptors, all of which can stymie the healing process.¹⁶ While it is not clear whether elevation of MMPs is responsible for delayed wound healing or merely the result of other changes occurring within the chronic wound, it has been shown that decreases in MMP-3 and MMP-9 are associated with wound healing.^{2,16,18,20,27} Thus, the decrease in protease levels observed in this study are consistent with beneficial changes in the wound environment.
The baseline (Day 0) levels of MMP-3 (~700 ± 100 ng/mL) in the wound fluids from the chronic wounds in this study were higher, as expected, than values reported in the literature for acute wound fluids (median levels of ~53 ng/mL for acute breast surgery wound fluid and ~274 ng/mL for acute colorectal wound fluid).²⁸ The baseline MMP-3 levels in this study with pressure ulcer patients in extended care and home health settings were statistically greater than those reported by Stechmiller et al¹² in a previous study involving pressure ulcer patients in an inpatient setting (360 ± 100 ng/mL; n = 6).¹² These differences suggest that the wounds in the current study had a biochemical profile less conducive to healing than the wounds in the previous inpatient study. With NPWT, the levels of MMP-3 in wound fluid in this study decreased to a level (~200 ng/mL) significantly lower than those observed in the previous inpatient study (~400–600 ng/mL; the levels were not significantly altered with NPWT). It is statistically highly suggestive (P = 0.054) that MMP-9 levels in this study (~6000 ± 2000 ng/mL; n = 8) were lower than the values reported for pressure ulcer fluids by Ladwig et al¹⁹ (MMP-9: 13000 ± 3000 ng/mL; n = 125).
TIMP-1, like the other 3 known TIMPs, is 21 to 29 kDa in size and inhibits all known secreted MMPs. An imbalance in the ratios of MMPs²⁴ and TIMPs has been associated with a myriad of pathological conditions, including chronic pressure ulcers.^{2,3,16,18,20,28} Decreased TIMP-1 expression is associated with fibroblast senescence, which may explain, in part, why the elderly are more likely to have chronic wounds.²⁹ TIMP-1 levels are also significantly lower—and sometimes even completely absent—in chronic wounds.^30,31 The levels of TIMP-1 at baseline (Day 0; ~700 ± 200 ng/mL) in the current study were higher than values reported by Ladwig et al¹⁹ (< 250 ng/mL) but lower than levels in a previous inpatient pressure ulcer study by Stechmiller et al¹² (1,500 ± 300 ng/mL). These results support the earlier postulate that the wounds in the current study had a biochemical profile less conducive to healing than the wounds in the previous Stechmiller inpatient study. These differences may be due to the level of debridement (> 24 hours prior to baseline), which may have been the most aggressive with the inpatients.³² Better nutritional support and offloading in an inpatient setting may also have contributed to the differences in these results.³³
It is important to note that the ratio of MMP to TIMP-1 may be more pertinent in providing diagnostic and prognostic value.¹⁹ The increased level of TIMP-1 in this study may be a reflection of tissues attempting to counter the significantly elevated levels of MMPs. MMP-3:TIMP-1 ratios in the current study decreased in a statistically significant manner from ~1.0 to ~0.4, which is statistically not different from ~0.5 reported by Stechmiller et al¹² for Day 0 through Day 7 in their inpatient study. This again corroborates the notion that the wounds in the current study may not have had biochemical profiles as favorable to healing as in the previous inpatient study.
TNF α and IL-1β are markers of inflammation.²⁰ TNF α is a pro-inflammatory mediator and typically participates early in the inflammatory phase of wound healing. Different cells, including activated macrophages and monocytes, B cells, T cells, and fibroblasts, secrete TNF α.³⁴ TNF α is expressed as a larger membrane protein that is then processed by a MMP (TNF α converting enzyme) into the biologically active trimeric form (52 kDa).³⁴ TNF α levels (~200 ± 100 pg/mL) in this study were consistent with the range (0 to 6,000 pg/mL) reported by Ladwig et al¹⁹ but lower than the values (800 ± 200 pg/mL) reported by Stechmiller et al¹² in their inpatient pressure ulcer study. This is reasonable, because the patients included in the Stechmiller et al¹² study were enrolled soon after being released from an intensive care unit and may have had increased levels of inflammation. Evaluation of serum levels in the Stechmiller study may have provided greater insight into the systemic levels of TNF α and offered an explanation for the higher concentration in wound fluid. IL-1β is an 18 kDa molecule that is secreted by a variety of cells, including immune, muscle, endothelial, and neural cells, and has pyrogenic qualities.³⁴ The baseline IL-1β values in this study (~5000 ± 2000 pg/mL) are similar to those reported for pressure ulcers in inpatients (10000 ± 4000 pg/mL) by Stechmiller et al¹² and those reported by Ladwig et al¹⁹ (0–12000 pg/mL). TNF α and IL-1β have been previously shown to stimulate the production of MMP-235 and MMP-336 in vitro. However, the current study showed decreases in MMP-3 levels but no changes in the 2 inflammatory cytokine levels that may reflect the complexity of interactions between different analytes that is absent in an in-vitro setting. It is possible that TNF α converting enzyme, a MMP that converts the precursor of TNF α to its active form, may have also decreased and contributed to the absence of change in concentration of TNF α from baseline (this MMP was not evaluated in this study).³⁴ Since there can be numerous redundant pathways involving other biomolecules for expression and/or activity of a particular protein, causative statements must be made with caution. To overcome the limitation of sample volumes, a proteomics approach in future studies would allow analysis of a much larger number of biomolecules and thereby allow exploration of other pathways.
The results from this study are also consistent with the postulate that mechanical strain created (or the relief thereof) at the interface between the foam and underlying tissue has a role in NPWT promoting wound healing.³⁷ Mudera et al³⁸ have shown that human dermal fibroblasts that are under strain but have not realigned themselves to accommodate this strain express increased levels of MMP-2 and MMP-3 compared with cells that accommodated the strain. In that same study, expression of TIMPs was shown to be less sensitive to mechanical stimuli.
Future assays of biomolecules should include evaluation of both tissue and wound fluid. Expression in tissue does not necessarily indicate secretion and vice versa. Examination of tissue and wound fluid would also help determine whether decreases in biomolecule concentrations may have been a result of dilution effects caused by more fluid being pulled through the tissue or that there was actual activation of cellular machinery. Study of the tissue receptors of these biomolecules would also be very important; although levels of a specific biomolecule may not have changed, changes in the quality and quantity of its receptors can have profound effects on local activity of that biomolecule.
The decrease in MMP-3, MMP-9, and MMP-3:TIMP-1 ratios are consistent with results from other studies that suggest NPWT helps normalize healing: 1) a human pressure ulcer wound fluid study¹² showed decreases in previously elevated levels of TNF α; 2) a non-survival pig study¹¹ exhibited early elevation in serum concentrations of IL-10, which is an anti-pro-inflammatory cytokine; 3) a rabbit study showed that application of NPWT normalizes immediate vascular response in injured tissue compared with tissue treated with saline-moistened gauze and is consistent with results from an ex-vivo swine study by Wackenfors et al.^39,40 The results from the current study may explain, in part, the success of NPWT in the treatment of wounds; however, more studies are needed to confirm this possibility. Results from this pilot study provide the data necessary to estimate sample size and help design and implement powered and more controlled studies in the future. Wounds to be included in future studies should be monitored for a specified length of time (eg, 2 to 4 weeks), be similar in location, and undergo similar treatment modalities prior to initiation of therapy to ensure balanced distribution of wound profiles across treatment groups. In these statistically powered studies with control arms, it would be important to monitor clinical outcomes, including clinical signs of inflammation, in conjunction with assaying of biomolecules.

Conclusion

In this pilot study, wound fluids from pressure ulcer patients in home health or extended care settings treated with NPWT had decreased levels of MMP-3, MMP-9, and MMP-3:TIMP-1 ratios during the initial week of treatment compared to baseline (P < 0.05). Future studies with control arms should be conducted to correlate the clinical effects of NPWT-driven reduction of these analytes.

Acknowledgments

The authors thank Mr. Oliver Grundmann for assistance with data collection and Ms. Charu Jain for quality audit of the data.