What is the Role of Point-Of-Care Fluorescence Imaging for Bacterial Load Identification

March - April 2023

London, United Kingdom

What is the Role of Point-Of-Care Fluorescence Imaging for Bacterial Load Identification in Diabetic Foot Ulcers?

Editorial Summary

We recently published 'Point-Of-Care Fluorescence Imaging Reveals Extent of Bacterial Load in Diabetic Foot Ulcers' in the International Wound Journal, which explores point-of-care fluorescence imaging using a prospective, single-blind, multi-center and cross-sectional clinical trial.1 In this editorial we discuss the relevance of this trial and the impact of new terminology: chronic inhibitory bacterial load (CIBL), and how it intends to help define the presence of bacteria at high loads that is distinct from infection, but may inhibit wound healing.

Introduction

Worldwide, there are almost half a billion people with diabetes. Of these, 1 in 3 will develop diabetic foot ulcer (DFU), which has an associated 1.89 increased mortality risk.1 The most important clinical complication is diabetic foot infection.2 In diabetic patients, 60% of foot ulcers will become infected during clinical management.3 Diabetic foot infections are associated with increased economic costs, in addition to significant morbidities. They are the most frequent diabetes related indication for hospital admission and the most common reason for lower extremity amputation in diabetic patients. Overall, long-term healing rates and outcomes in these patients are poor.2

In addition to infection, other local factors influencing the healing rate of ulcers are pressure at the wound site and adequate blood supply to the site. When treating ulcers, it is often difficult to establish whether the wound is infected or not. High bacterial loads are often found in ulcers that do not appear clinically remarkable.4 Therefore, another important factor to consider in determining the rate of ulcer healing is bacterial infiltration of the wound, with or without overt clinical infection. The presence of a high burden of bacteria at the site of a wound has previously been shown to disrupt wound healing. Left unaddressed, bacterial endotoxins released from Gram-negative bacteria e.g., Pseudomonas aeruginosa, attract immune cells that perpetuate the inflammatory response, stall the healing cascade and force the wound to remain chronically open.5,6 Polymicrobial biofilms, present in 68% - 100% of DFUs, similarly contribute to wound chronicity by stimulating inflammation.8,9 The degree of bacterial infiltration varies and is dependent upon bacterial virulence and species.4,7,8

Cambridge MA, United States

The presence of high numbers of bacteria and formation of biofilms at the site of DFUs perpetuates the chronic wound cycle and increases the risk of infection. Earlier and more accurate detection of bacteria, with appropriate management, could reduce the risk of associated morbidity.1

Dr David G. Armstrong

Department of Surgery, Keck School of Medicine of University of Southern California

Los Angeles CA, United States

Dr Thomas E. Serena SerenaGroup® Research Foundation

Dr Michael E. Edmonds

Diabetic Foot Clinic, King’s College Hospital Foundation Trust

Chronic inhibitory bacterial load (CIBL) is defined as:

“The chronic presence of bacterial microorganisms in a wound or its surrounding tissue at loads which can damage tissues and be inhibitory to healing, as well as require clinical intervention, with or without the presence of clinical symptoms.”

Point-of-Care Fluorescence Imaging

There are few reliable diagnostic tools to directly detect the presence of bacteria in wounds. Of those available, most do not provide immediate results. Point-of-care fluorescence imaging (FL-imaging) of bacteria is a biotechnology well positioned to address this clinical concern.

Under current guidelines, sampling and microbial analysis of a wound is done only if clinical infection is suspected.2 Although this allows treatment with appropriate antibiotics, superficial sampling techniques are suboptimal and are impeded by the presence of bacterial biofilms.1 This also misses the opportunity to treat patients with subclinical infection, where high levels of bacteria are present at the wound site, delaying healing and increasing the risk of future clinical infection.

Point-of-care FL-imaging uses endogenous FL signals, produced by bacterial metabolites and virulence factors to identify bacterial loads in wounds. It is objective, sensitive and noninvasive. FL-imaging is able to identify most bacterial species at clinically relevant levels, considered to be >104 colony-forming unit per gram (CFU/g), and crucially can detect biofilmencased bacteria.1

Since many patients who do not have clinically overt infection do have high levels of bacteria present at the wound site, the term infection is not particularly helpful in characterizing the degree of bacterial infiltration. This value is clinically important, as the presence of high levels of bacteria may inhibit wound healing, even in the absence of infection. Therefore, a new terminology, chronic inhibitory bacterial load (CIBL), helps to make the distinction between infection and high levels of bacterial infiltration requiring intervention.

The FLAAG Clinical Trial

In summary, the FL-Imaging Assessment and Guidance (FLAAG) clinical trial was a prospective, single-blind, multi-center, and cross-sectional study. It included 350 adults (>18 years) presenting with wounds of unknown infection status.10 Patients were recruited from 14 outpatient wound care centres across the United States between May 2018 and April 2019. An independent third party (Ironstone Product Development, Toronto, ON, Canada) was used to control for bias and to ensure appropriate blinding. The study aim was to compare the performance of standard clinical signs and symptoms (CSS) assessment using International Wound Infection Institute (IWII) guidelines, to CSS in combination with FLimaging, to detect clinically relevant bacterial loads. The clinicians received relevant training prior to commencing the study. They reviewed the patient’s history and inspected the wounds for all signs and symptoms of covert, overt, and spreading infection, identified by the IWII 2016 guidelines.11 Each sign and symptom was recorded when detected, including delayed healing beyond expectations; this was identified if the wound area had not reduced by at least 30% during the prior 4 weeks of care.

Post-hoc analysis of the FLAAG was conducted. The recorded CSS was reviewed to identify wounds that fulfilled the International Working Group of the Diabetic Foot (IWGDF) criteria for infection. Immediately following clinical assessment, standard and FL images were taken using the FL-imaging device (MolecuLight i:X, Toronto, Canada). This advanced imaging technology creates a map of high bacterial loads in and around wounds, without the use of contrast agents.10,12 Clinicians participating in the trial underwent didactic and hands-on training on use of the device; they were also trained on image interpretation and were required to pass an image interpretation certification test with a score of >80%.

Point of care testing (POCT) is defined as diagnostic testing that is performed at or near to the site of the patient with the result leading to a potential change in the care of that patient.

Red FL on images indicates the presence of endogenously produced porphyrins from most common wound pathogens, while cyan FL indicates pyoverdine virulence factors from Pseudomonas aeruginosa specifically, both at loads >104 CFU/g.13,14

Up to 3 punch biopsies were collected from each DFU after cleansing of the wound with saline and gauze. More than 1 biopsy was only taken if an area of interest was identified or red or cyan FL was detected outside the wound centre.

Quantitative culture was subsequently performed, ensuring optimal conditions for difficult-to-culture bacteria were maintained.10 Diluted biopsy samples were cultured on various agars in conditions to support both aerobic and anaerobic growth.10,15 To identify specific bacterial species, matrix assisted laser desorption ionisation-time of flight mass spectrometry (Bruker Daltonics) was used.

Statistical analysis of data calculating both sensitivity and specificity at various bacterial thresholds was performed using MedCalc© Version 19.1.5. This was calculated with 95% confidence intervals (exact Clopper-Pearson). Bacterial loads were compared between the wound centre and periwound biopsies using a 2-sided paired student t-test.1

Results

Data regarding participant characteristics was noted. Importantly, participants with skin tones from across the Fitzpatrick scale (I to VI) were included. Most had a DFU of >3 months duration.1

In this study, 138 DFUs were examined; the majority of these (131) had some bacterial presence. Most ulcers that had some bacterial presence were also found to have high bacterial loads. Of the 131 ulcers where bacteria were found, only 6.1% (8/131) had bacterial loads below 104 CFU/g, while 93.9% (123/131) had bacterial loads exceeding 104 CFU/g, and 83.2% (109/131) had bacterial loads exceeding 105 CFU/g.1 The average bacterial load of DFUs with confirmed bacterial presence was 1.44 x 108 CFU/g. The number of bacterial species varied between wounds, with some ulcers having as many as 8 different species. The most common species found was Staphylococcus

aureus (52.4%) and the average number of bacterial species per biopsy was 2.74.1

Although red and/ or cyan FL was detected in the majority of DFUs, as the bacterial load increased so did the proportion of DFUs with FL indicating bacterial loads. Red or cyan FL was identified in 92.3% of DFUs with the highest bacterial loads. In wounds where high volumes of bacteria were identified, IWGDF infection criteria were largely missing.

Five CSS were included in the present study, following those in the IWGDF criteria; these were swelling, erythema, warmth, and purulent discharge. In ulcers where there were very high levels of bacteria (>108 CFU/g), the most common IWGDF criteria detected were swelling (11.5%), erythema (15.4%), pain (11.5%), and local warmth (11.5%). Where bacterial levels were very low, only swelling was detected. Purulent discharge was consistently the least commonly observed. Of the clinically assessed signs and symptoms, none proved better than chance at predicting high bacterial loads of greater than 104 CFU/g.

Sensitivity did not improve at higher bacterial loads for 3 out of 5 IWGDF criteria. The only criteria which did show a high sensitivity at higher bacterial loads were pain and purulent discharge, however their prevalence was low. The low sensitivity and 1-specificity values indicate a poor predictive value of all 5 IWGDF criteria when assessing high bacterial load.1

Table 1: Sensitivity in detection of bacterial load using CSS according to the IWGDF criteria, compared to CSS with FL-imaging.

In the study, the most commonly observed clinical sign was delayed wound healing beyond expectations. Generally, as bacterial load increased, the time taken for the wound to heal also increased. In over 50% of cases (52%), DFUs with bacterial loads of >104 CFU/g were healing delayed. This increased to 64.7% of wounds with bacterial loads of 107 - 108 CFU/g. Most wounds demonstrating a delay in healing were FL positive (70.0% - 95.5%).1

Sensitivity of delayed healing beyond expectations for predicting high bacterial loads was high across all bacterial thresholds, but specificity was lower than IWGDF criteria.1

In this study, 84.2% of the DFUs that exhibited red or cyan bacterial FL had FL indicating bacteria outside of the wound bed; this was mostly confined to callused tissue in the periwound (2cm radius extending out from the wound edge).16 Red FL signals within this region usually appeared blush pink to yellow due to the bacteria being below the surface. Bright red FL signals observed were attributed to the presence of higher bacterial loads at or near the callus surface.

If an additional biopsy was taken, the number of species and bacterial load at the wound centre was compared to the biopsy from outside the wound bed. The mean bacterial load from regions of red or cyan FL in the periwound region was significantly higher than the mean bacterial load of biopsies collected from the wound centre, implying that higher bacterial loads are present at the edge of the wound.1

Chronic Inhibitory Bacterial Load

The findings of this study led to a proposal of new clinical terminology: chronic inhibitory bacterial load (CIBL). This term refers to the state of bacterial infiltration within a wound that resists healing, but the process is not linked to any known clinically perceptible sign or symptom.1 CIBL is a state that is both CSS independent and is characterized by an inhibition of normal healing processes.

The DFUs in this study experiencing delayed healing were likely to be FL-positive, regardless of the bacterial threshold. Most DFUs that were FL-positive had signals indicating bacteria in the peri-wound region (84.2%). In addition, of the clinically assessed signs and symptoms,

none were better than chance at predicting high bacterial loads.1 These findings suggest that a state exists where there are high bacterial loads, resulting in reduced wound healing, without any clinically apparent presentation. Introduction of this new terminology seems to help address this diagnostic gap.

Clinical Implications

In this study, CSS did not correlate with high bacterial loads. In fact, CSS were largely missing, and all 5 IWGDF criteria were infrequent. This suggests that clinical assessment alone is not a suitable method of detecting high bacterial loads, as there is frequently an absence of signs and symptoms even when there are many bacteria present. The term ‘CIBL’ illustrates this important pathogenic state. There is no connection between CIBL and any bacterial threshold, it merely describes the state in which a persistently elevated bacterial load results in pathology, including delayed healing. There is also no link to infection status of the wound.

CIBL may exist in a wound at any point beyond contamination, regardless of clinical markers of infection; it can therefore be addressed with bacteria-based wound care, which may include debridement and cleansing.1,11 This concept acknowledges the role of the biofilm in wound chronicity and is also suited to holistic wound infection prevention and management.

In clinical practice, CIBL can be utilized as an indicator for intervention prior to the onset of infection. Current clinical guidelines do not respond to asymptomatic but potentially pathogenic bacterial loads, despite acknowledging the value of their early and proactive management. This may be due to the inability to clinically detect them. Fl-imaging overcomes this diagnostic limitation, as it can visualize high bacterial loads across the entire wound and periwound.

Previous evidence has demonstrated high positive predictive values (PPV) of FL-imaging, and sensitivity greater than that of the CSS of infection. This study is consistent with previous findings, and has also demonstrated that while the sensitivity of CSS is not proportional to bacterial load, there is a proportional increase in sensitivity of FL-imaging.1 Furthermore, DFUs often present with macerated tissue. This allows futher bacterial colonization,

What is the Role of Point-Of-Care Fluorescence Imaging for Bacterial Load Identification in Diabetic Foot Ulcers?

particularly of organisms like Pseudomonas aeruginosa and Staphylococcus aureus

FL-imaging provides critically important information on the distribution of CIBL. Imaging in this study demonstrated that 84.2% of DFUs contain elevated bacterial loads in the periwound region and in callus tissue. Clinically, this might indicate that aggressive wound hygiene strategies like sharp debridement may be warranted when CIBL is identified in the DFU periwound, given the depth to which bacteria can extend. In future, the periwound should be recognized as a region frequently colonized with subsurface bacterial loads.1

Strengths and Limitations

One of the main strengths of the study was its generalizability to the overall DFU population. Minimal participant exclusion criteria were enforced, leaving a heterogenous sample of patients. Recruitment was also wide, from numerous clinical sites, and the clinicians were from a range of wound specialties. The diagnostic accuracy of the FL-imaging was ensured by completing quantitative culture analysis to confirm bacterial loads.1

The major limitation of the study was the lack of experience of the clinicians with the technology, which may have influenced sensitivity data. There are also other systemic contributors which influence wound healing which were not considered here.

Conclusion

Current practice does not include a strategy for the treatment of high bacterial loads that are asymptomatic. This presents a clinical problem, as high bacterial presence at the wound site has been associated with ongoing wound chronicity and increased infection risk.

The lack of sufficient and reliable means of identifying the presence of high bacterial loads in asymptomatic patients has resulted in a gap in understanding of the associated clinical implications. FL-imaging provides a solution to this; by facilitating appropriate identification of bacterial burden, there is a potential to aid early intervention and monitor treatment effectiveness, during and after local wound treatments. There is also a possible means of more effectual antimicrobial stewardship, to limit antibiotic and antimicrobial dressing

prescriptions and to improve wound healing outcomes.

The prevalence and distribution of bacterial burden in DFUs has historically been underestimated, and its pathogenicity is unclear. There has been an absence of a clinical definition for such a finding in asymptomatic patients. It is anticipated that the definition of CIBL will spark a paradigm shift in DFU wound assessment and management, and it is hoped that a clearer clinical definition of bacterial burden in patients, where typical symptoms of infection are absent, will encourage earlier intervention along the bacterial infection continuum, thereby preventing sequelae of infection and supporting improved DFU outcomes.

References

1. Armstrong DG, Edmonds ME, Serena TE. Point-ofcare fluorescence imaging reveals extent of bacterial load in diabetic foot ulcers. Int Wound J. 2023;20(2):554-66.

2. Lipsky BA, Senneville É, Abbas ZG, Aragón-Sánchez J, Diggle M, Embil JM, et al. Guidelines on the diagnosis and treatment of foot infection in persons with diabetes (IWGDF 2019 update). Diabetes Metab Res Rev. 2020;36 Suppl 1:e3280.

3. Jia L, Parker CN, Parker TJ, Kinnear EM, Derhy PH, Alvarado AM, et al. Incidence and risk factors for developing infection in patients presenting with uninfected diabetic foot ulcers. PLoS One. 2017;12(5):e0177916.

4. Xu L, McLennan SV, Lo L, Natfaji A, Bolton T, Liu Y, et al. Bacterial load predicts healing rate in neuropathic diabetic foot ulcers. Diabetes Care. 2007;30(2):378-80.

5. Rippon MG, Westgate S, Rogers AA. Implications of endotoxins in wound healing: a narrative review. J Wound Care. 2022;31(5):380-92.

6. Phillipson M, Kubes P. The Healing Power of Neutrophils. Trends Immunol. 2019;40(7):635-

47.

7. Lantis JC, 2nd, Marston WA, Farber A, Kirsner RS, Zhang Y, Lee TD, et al. The influence of patient and wound variables on healing of venous leg ulcers in a randomized controlled trial of growth-arrested allogeneic keratinocytes and fibroblasts. J Vasc Surg. 2013;58(2):433-9.

8. Goldberg SR, Diegelmann RF. What Makes Wounds Chronic. Surg Clin North Am. 2020;100(4):681-93.

9. Johani K, Malone M, Jensen S, Gosbell I, Dickson H, Hu H, et al. Microscopy visualisation confirms multi-species biofilms are ubiquitous in diabetic foot ulcers. International Wound Journal. 2017;14(6):1160-9.

10. Le L, Baer M, Briggs P, Bullock N, Cole W, DiMarco D, et al. Diagnostic Accuracy of Pointof-Care Fluorescence Imaging for the Detection of Bacterial Burden in Wounds: Results from the 350-Patient Fluorescence Imaging Assessment and Guidance Trial. Adv Wound Care (New Rochelle). 2021;10(3):123-36.

11. Wounds International. International Consensus Update 2016 International Wound Infection Institute (IWII) Wound Infection in Clinical Practice: Principles of best practice. 2016 [cited June 22, 2022]. Available from: https://www.woundsinternational.com/resources/details/woundinfection-inclinical-practice-principles-of-bestpractice

12. Rennie MY, Dunham D, Lindvere-Teene L, Raizman R, Hill R, Linden R. Understanding Real-Time Fluorescence Signals from Bacteria and Wound Tissues Observed with the MolecuLight i:X(TM). Diagnostics (Basel). 2019;9(1).

13. Raizman R, Little W, Smith AC. Rapid Diagnosis of Pseudomonas aeruginosa in Wounds with Point-Of-Care Fluorescence Imaing. Diagnostics (Basel). 2021;11(2).

14. Jones LM, Dunham D, Rennie MY, Kirman J, Lopez AJ, Keim KC, et al. In vitro detection of porphyrin-producing wound bacteria with real-time fluorescence imaging. Future Microbiol. 2020;15:319-32.

15. Serena TE, Bowler PG, Schultz GS, D’Souza A, Rennie MY. Are Semi-Quantitative Clinical Cultures Inadequate? Comparison to Quantitative Analysis of 1053 Bacterial Isolates from 350 Wounds. Diagnostics (Basel). 2021;11(7).

16. Andersen CA, McLeod K, Steffan R. Diagnosis and treatment of the invasive extension of bacteria (cellulitis) from chronic wounds utilising point-of-care fluorescence imaging. International Wound Journal.

2022;19(5):996-1008.