INTRODUCTION
Malignant fungating wounds (MFWs) afflict 5% to 14% of the 609,640 patients with advanced cancer in the United States.1-6 An MFW is a nonhealing wound resulting from aggressive proliferation of malignant cells and tumors that infiltrate the epidermis, blood and lymph vessels, and underlying structures in patients with advanced cancer.1,5,7 Unchecked local tumor extension inflicts widespread tissue damage and causes disfigurement, loss of vascularity with subsequent tissue hypoxia and necrosis, polymicrobial proliferation, and fungating ulceration of the wound bed.5,7,8 These wounds are visible markers of underlying neoplastic disease, predominantly developing during the last 6 months of life and indicate impending end of life.1,8 Given the incurable nature of MFWs, palliative strategies to alleviate symptoms are imperative to improve patients' physical, emotional, social, and functional well-being as well as overall quality of life until the end of life.
Malignant fungating wounds elicit distressing symptoms (such as pain, odor, exudate, bleeding, and pruritus)7 and exert a devastating impact on the physical, psychological, and functional health of patients in the last months of life.1,9-12 Malignant fungating wound symptoms are postulated to be influenced by their microbiome.13-17 Malignant wound beds are reported to be polymicrobial, but their microbiome has not been fully examined.13,14,16 Understanding the role of the microbiota in developing or amplifying the severity of symptoms associated with MFW will help to establish a better understanding of microbial virulence and develop targeted palliative wound care interventions to decrease symptoms of MFWs and alleviate suffering.
Prevalence and Incidence
Malignant fungating wounds may arise from any type of underlying malignancy; the most prevalent are associated with breast cancer (66%), followed by head and neck tumors (24%).4,8 Malignancies of the groin, genitals, and back combined account for 3%, and all other sites account for the remaining 8%. Local extension of primary breast tumors is the most common cause of fungating lesions in women while cutaneous metastasis from lung cancers is most often seen in men.5,8 This overwhelming (49%-66%) prevalence of fungating breast wounds explains the disproportionate number of women affected by MFWs; furthermore, increased cancer occurrence with age explains the disparate incidence of MFW among those older than 50 years.6
Physiology
Malignant fungating wounds originate from 1 of 3 etiologies: primary skin neoplasms, local extension and integumentary erosion from primary tumors or malignancy recurrence, and from metastatic cutaneous lesions.1,5,7 Metastatic invasion of the basement membrane is essential for the development of an MFW. These wounds tend to expand rapidly; occur along pathways of least resistance such as surgical incisions, tissue planes, and blood or lymph vessels; and result in lymphedema due to local lymphatic vessel invasion.5,8,11
Initially, MFWs present as inflammation, peau d'orange appearance, smooth raised nodules varying in size and pigmentation, erythematous plaques or patches, areas of alopecia, or violaceous papules.12 They rapidly progress to cauliflower-shaped lesions (proliferative) or an ulcerated area (erosive), sometimes forming fistulas or a sinus tract.5,12 Morphology of MFWs is unique compared to other wounds: fungating lesions are protruding and grotesque, while ulcerating lesions are cavity-forming and prone to bleeding, infection, and malodorous exudate.11
The wound-associated microbial communities found in necrotic tissue include anaerobic bacteria, such as Bacteroides and Clostridium species, which create malodor by producing the volatile short-chain organic acids (n-butyric, n-valeric, n-caproic, n-heptanoic, and n-caprylic).12,15,18 These fatty acids combine with the amines and diamines (cadaverine and putrescine) produced by the proteolytic enzymes of other bacteria to create pungent odors that provoke the gag reflex.19 Importantly, fatty acid volatiles are commonly associated with the anaerobes that thrive in necrotic tissue, a common component of MFWs.12 In an earlier morphoqualitative analysis of MFWs, Tamai and colleagues20 reported a significantly higher percentage of necrotic tissue, a known medium for anaerobe growth and exudate production, in patients with MFW and moisture-associated skin damage (MASD) group compared to a group free from MASD (P = .066).
Microbiome
The human microbiome is a collection of microorganisms living within the human body.13,14 A microbiome (ie, the bacterial community structure) of MFWs has also been identified.13-15 Early studies show that malignant wound beds are polymicrobial, with a predominance of Staphylococcus.16,17 Thus, the bacterial community structures of the MFW may influence the development and severity of wound symptoms, such as pain, odor, exudate, bleeding, and pruritus.16-20
The purpose of this systematic review was to evaluate evidence regarding the relationship between microbiome and symptoms of MFWs. Specific aims were to investigate evidence regarding: (1) the microbiome and metabolome of MFWs, and (2) the relationships between wound microbiota and symptoms of MFWs.
METHODS
We aimed to evaluate research articles published in peer-reviewed and English-language journals between January 1, 1995, and January 1, 2020; we selected these dates because it contains the majority of research on MFWs. Prior to conducting the literature search, we defined key microbiome research concepts: microbiota, microbiome, and metabolome (Box).21 Symptoms were defined as subjectively perceived indicators of abnormal biological or physiological changes that may or may not be observed objectively.22-25
BOX.Key Concepts in Microbiome Researcha
Microbiota: The assemblage of the microorganisms present in a defined environment
Microbiome: The entire habitat, including the microorganisms, their genomes, and the surrounding environmental conditions
Metabolome: The census of all the metabolites present in any given strain or single tissue
aAdapted from Marchesi and Ravel.21
Search Strategy
The review was conducted in 2 phases after consultation with a health librarian to ensure reliability and accuracy of Medical Subject Headings (MeSH). The first author performed the initial search using key search terms "malignant fungating wound" or "malignancy" or "neoplasm" or "malignant wound" and "microbiome" or "microbiota" or "metabolome" or "bioburden" or "bacteria." Additionally, an ancestry search of retrieved articles was performed to identify additional studies.26 Using various combinations of the key terms, the following databases were searched: Excerpta Medical Database (EMBASE), PubMed, the Cumulative Index for Nursing and Allied Health literature (CINAHL), and Cochrane Library. The Figure is a PRISMA flow diagram of our search; our initial search retrieved 12,442 articles; 12,439 were retrieved by searching electronic databases and 3 were found based on ancestry search. After removal of 4574 duplicates, 7868 articles remained (Figure).
Inclusion and Exclusion Criteria
Inclusion criteria for this review were: (1) quantitative (observational) or interventional research on MFWs, and (2) quantitative (observational) research on the microbiome or metabolome of MFWs. Exclusion criteria were: (1) research that focused on other types of wounds (ie, pressure injuries and Marjolin's ulcers); (2) interventional research that did not include microbiome or metabolome data; (3) individual case studies; (4) unpublished dissertations and abstracts26; and (5) non-English language publications.
Article Screening and Selection
There were 4 steps in the screening and selection process: (a) a title review, (b) an abstract review, (c) a full article review, and (d) an evaluation of the quality of articles (see the Figure). We screened 7868 titles and identified 724 articles; then, we screened 724 abstracts and identified 16 abstracts that were read in full. A review of the 16 articles identified 7 that met inclusion criteria. Nine studies were excluded: 6 did not focus on the microbiome or metabolome and 3 focused primarily on interventions without microbiota data. An adapted quality assessment tool using a 14-item index was used to evaluate the quality of the 4 quantitative observational and 3 interventional studies identified.26-29 For this systematic review, studies that received an affirmative score of at least 10 out of 14 were considered to have adequate quality.27,28 All 7 articles were deemed adequate and included in this review (Table 1).
Data Extraction and Analysis
Detailed critical appraisal of each included quantitative study was performed using an adapted evaluation tool that quantitatively assessed overall quality of each study.27 We extracted descriptive data, key results (technological approaches, microbiome data, and metabolome data), and evaluation of risk of bias in terms of study weaknesses (Table 2).26,28 The heterogeneity of the quantitative studies and the use of a variety of psychometric tools and bacteriological procedures limited our ability to perform a formal meta-analysis of pooled findings.
RESULTS
This systematic review identified 4 quantitative observational studies16,18,20,30 and 3 quantitative interventional studies.17,31,32 The overall quality of the 7 studies was adequate (mean = 11.42 +/- 1.49, range 10-14) (Table 2).
Four studies examined the microbiome of MFW,16,17,31,32 1 investigated the metabolome,18 and 2 studies examined both microbiome and metabolome.20,30 Characteristics of the combined sample were heterogeneous and sampling methods and criteria were not reported in 3 studies.18,30,31 In addition, 4 studies did not report attrition rates.16,18,31,32 The pooled population was 193 (mean sample size = 27.57 +/- 21.17, range 5-67). The majority of the participants were women (n = 183) and older adults (>50 years). Only 1 of the 5 studies that reported age included subjects younger than 50 years.31 Ninety-two percent of participants in the pooled sample had MFW arising from breast cancers (n = 177).16-18,20,30-32 Most of the remaining sample had wounds located in the head and neck region (n = 10),17,18 with the remaining 6 characterized as "other."17
Bacteriological Procedures and Technological Approaches
Procedures to characterize microbial communities varied in the reviewed studies. Table 2 outlines the bacteriological procedures and technological approaches used in each study. Five studies used traditional culture-dependent approaches.16,17,20,31,32 All 5 microbiome studies used unspecified curettage material to sample various tissues and exudates, but swab sample techniques were only partially described. These cultivation-based studies used disparate sampling procedures; 2 collected fresh exudate,17,32 2 pooled (old) and fresh exudate,20,31 and 1 did not discuss the origin of the substance swabbed.16 Biofilm analysis was conducted by epifluorescence microscopy.16
Metabolomic technological approaches included gas chromatography-mass spectrometry-olfactometry, which was used in 2 studies to determine the chemical identity of odorants and classify odor intensity.18,30 Current metabolomics procedures were also limited. Metabolites were identified in 1 cross-sectional, observational study by high-performance liquid chromatography analysis, gelatin zymography, and skin pH meters.20
Microbiome
All studies found malignant wound beds to be polymicrobial: 20 different species of aerobes and 14 species of anaerobes were identified (see Supplemental Digital Content 1, available at: http://links.lww.com/JWOCN/A59). One study detected 25 different species of bacteria and wound colonization, with a median of 2 colonies per patient.17 In this study, combined qualitative cultures demonstrated a predominance of aerobes from the phyla Firmicutes, Actinobacteria, and Proteobacteria, and anaerobes from the phyla Proteobacteria, Bacteroidetes, Firmicutes, and Fusobacteria.17Staphylococcus was the predominant genus in 2 studies.16,17 The number of distinct species detected ranged from 6 to 54,16,30 with a median of 2 to 6 species colonizing each patient's wound.17
Biofilms are extracellular polymeric substances that facilitate microbial adherence to the wound bed and provide a medium for chemical signaling and pathogenesis.33,34 Biofilms were found in 35% of MFWs in 1 study, but were not associated with specific bacteria.16 The remaining studies did not include specific techniques to detect biofilms or report data about biofilms.17,18,20,30,31 Presence of drug-resistant bacterial species was not specifically reported in 5 studies.18,20,30-32 Two studies did report methicillin-resistant Staphylococcus aureus, although other multi-drug-resistant strains were not identified.16,17
The bacteriological outcomes from 2 of the 3 interventional studies focused on quantitatively and qualitatively evaluating how topical interventions altered the microbial microenvironment,17,31 but they did not use a single, standard approach to examining or reporting the outcomes. Neither honey nor silver dressings significantly altered the microbiome.17 The antibacterial effects of the 2 dressings could not be confirmed quantitatively; 69% of wounds contained the same species at baseline and at the end of the study (honey dressings P = .60, silver dressings P = .26).17 Aerobic and facultative bacteria remained unchanged in all 5 participants after a median treatment length of 37 days with metronidazole gel; however, in 4 patients anaerobic colonies disappeared or decreased.31
Metabolome
Two observational studies evaluated the metabolome of MFWs using gas chromatography-mass spectrometry-olfactometry to determine the source of malodor.18,30 Cancer wound-derived odor was associated with dimethyl trisulfide (DMTS) and 4 fatty acid volatiles: acetic acid (sour odor), isobutyric acid (cheese odor), butyric acid (cheese and vomit odor), and isovaleric acid (cheese and foot odor).18 The source of DMTS was not attributed to a specific bacterium. Proteus mirabilis and Fusobacterium necrophorum produced the strongest and most typical malignant wound odor.30 Volatile organic compounds collected from primary dressings were found to contain bacterial metabolites associated with the most typical MFW odor: dimethyl disulfide, DMTS, phenol, indole, and 3-methylbutanal.18,30
The relationship between periwound MASD and the activity of various candidate-irritating factors was analyzed through pooled exudates using high-performance liquid chromatography. In addition, this relationship was analyzed using fresh exudate via gelatin zymography, and measuring exudate and periwound skin pH with a skin pH meter.20 Putrescine levels were higher in the exudates of participants with MASD compared to those from the non-MASD group (P = .008), and cadaverine was only present in the MASD group (P = .016).20 The study did not find significant differences in exudate pH (P = .125). Furthermore, the enzymatic activity of 2 different matrix metalloproteinases (MMP-2 and MMP-9) in the wound bed exudate was not significantly different between groups (P = .963).20
The Association Between Microbiome, Metabolome, and Symptoms
The association between symptom occurrence and characteristics (timing, intensity, quality, and distress)25 and MFW microbiota was examined using patient-reported and clinician-reported symptom outcomes,16,30-32 phenotypical wound status assessment data,16,17,20,32 or a combination of both16,32 in conjunction with various laboratory techniques examining the microbiome and metabolome (Table 3). Two studies reported phenotypical wound characteristics but no symptom outcomes.17,20 Five studies focused on either a single symptom18,30,31 or multiple symptoms (pain, odor, exudates, and bleeding),16,32 but none examined all 8 symptoms associated with MFWs.35
Microbiome and Symptom Burden
Technological approaches and bacteriological procedures varied among the 7 studies (Table 2). The bacterial spectrum found in MFWs significantly influenced symptom occurence.16 Symptom occurrence increased with an increase in number of bacterial species (P = .0003) and the presence of at least 1 anaerobe (P = .0006) in malignant wound beds.
Two observational studies examined wound pain as a symptom outcome.16,32 An evaluation of patient-reported wound pain intensity and wound pain control at baseline, day 21 and day 42, found little change in pain ratings over time despite the type of dressing or pain management regimen.32 Pain severity was graded by clinicians on a scale of 1 to 3 (grade 1 required nonopioid analgesia, grade 2 required light opioids, grade 3 required opioids).32 Grading criteria and differentiation between "light opioids" and "opioids" were not provided. Detailed reporting of systemic and topical analgesic management was also not provided.32 In another study, patients reported pain was significantly higher in the presence of >=105/g bacteria (P = .04); however, biofilms were not significantly associated with pain (P = .58).16
An interventional study of 5 women with breast MFWs reported that metronidazole gel applied to a malodorous MFW bed decreased or eliminated anerobic colonies in 1 to 5 days, resulting in patient- and clinician-reported disappearance of odor.31 A threshold concentration of >104/g was found to be predict the emergence of odors (P = .02).30 Wounds with strong odors, as reported by patients and clinicians between baseline and day 21, treated with systemic metronidazole demonstrated an overall reduction in wound odor intensity. The presence of a biofilm did not influence odor occurrence or severity (P = .88).16
Biofilms did not influence the occurrence or severity of exudates (P = .64).16 A study of exudate showed that putrescine levels in MFWs were higher in a group with MASD compared to a group free from MASD (P = .008).20 In addition, the researchers found that cadaverine was present only in the group with MFW and MASD (P = .016).20
Biofilms did not significantly increase the occurrence of bleeding.16 Instead, biofilm presence was associated with a decreased risk of provoked bleeding (P = .06).16 Local wound care controlled bleeding, although there was no association identified between bleeding and microbiome.16
Phenotypic Correlations to Microbiome and Metabolome
In a study of 67 breast cancer patients, Pseudomonas was found less often in wounds that decreased in size (P = .089 and P = .013 for honey and silver dressings, respectively).17 However, this difference was not statistically significant and no other pathogen showed a significant difference in prevalence between wounds that did or did not decrease in size.17 With no significant association between an increase in wound size over time and microbiota, a trend toward cavitation of the wound over time and an absence of color variation was found; most tissue was yellow (fibrin slough, tumor necrosis) and red (tumor buds).16 Periwound skin showed signs of inflammation in 52% of participants; however, this inflammation seemed to reflect the tumor mass, rather than infection. Similar findings in another study demonstrated that wound surface area remained globally stable: 50% of the wounds in this study had >=40% yellow (fibrin or soft necrosis/slough) and red tissue (budding malignant tissue) that remained stable throughout the study; 74% of the wounds had periwound inflammation secondary to tumor burden.32
DISCUSSION
We systematically reviewed studies of MFW and found that their microbiome was polymicrobial, with a predominance of aerobes, and a proliferation of Staphyloccocus.16,17 The distinct bacterial species detected ranged from 6 to 54, and aerobes and anaerobes from 7 phyla were identified.16,17,30 The bacterial species from MFW biomes included the Gram-positive Firmicutes and Actinobacteria, along with Gram-negative Proteobacteria and Bacteroidetes, which are a common feature of skin microbiome.15 Bacterial biofilms can cause wound infections and show resistance against antibiotics.33,34 Although methods to detect biofilms and/or multi-drug-resistant bacterial strains were not consistently used in the studies we retrieved, one found that 35% of MFWs had biofilms that were not associated with a specific bacterial species.16 Finally, while the diversity of skin microbiota varies by skin site (sebaceous vs moist or dry sites), comparison of the microbiome characteristics of malignant wounds from different skin sites/body regions has not been examined.15
Limitations and Strengths of the Literature
The major limitations of the current literature included the relatively small sample sizes, the use of various, biased, culture-dependent bacteriological procedures and different technological methods to examine the microbiome and metabolome of MFWs. The culture-based diagnostics used in the studies make it difficult to differentiate benign wound colonization from bacteria with problematic bioburden; to identify the full spectrum of bacterial species in wounds; and fail to identify metabolic components.13 Additionally, the inconsistent use of different instruments, lack of validity and reliability testing to assess MFW symptom severity, and symptom control limited our ability to conduct a formal meta-analysis.
Despite these limitations review findings provide preliminary evidence suggesting an association between MFW microbiota and wound symptoms. The amount of bacterial bioburden may be associated with occurrence of pain, and type of flora. Anaerobic bacterial species were found to be associated with odor occurrence and severity, along with exudate. In the light of these findings, future research should focus on further examining the association between MFW symptom occurrence and MFW microbiome and metabolome. This review also provided initial evidence that the microbiome did not affect wound size; however, devitalized tissue and subsequent exudate production were associated with periwound MASD.20
Given the paucity of evidence identified in this review, our understanding of the microbiome and metabolome of malignant wounds and their association with symptom occurrence and severity remains limited. The heterogeneity in sample characteristics, cancer etiologies, and wound locations in the studies led to risk of bias due to heterogenous host/wound environment, ultimately confounding efforts to definitively associate specific aspects of the microbiome with clinical phenotypes and wound symptoms.
Recommendations for Practice
* Topical antimicrobial interventions may reduce bioburden of MFWs and decrease symptom distress.
* Reduction or elimination of necrotic tissue may decrease odor and exudate severity.
* Periwound skin barriers are recommended to prevent periwound MASD.
CONCLUSIONS
We found limited evidence suggesting that the presence of putrescine in exuadate,20 the presence of DMTS and fatty acids in exudate,18 the number of bacterial species, a high number of bacteria, and/or the presence of anaerobes may be contributing factors to symptom occurrence and severity.16,30,31 To advance scientific knowledge, research is urgently needed to comprehensively examine the microbiome and metabolome of MFWs using modern genomic technology, and develop a psychometric tool that comprehensively measures all symptoms of MFWs. Comprehensive examination of the microbiome and metabolome of MFWs and their association with symptom characteristics is the first step to identify targeted systemic and topical microbicides and therapeutics.
KEY POINTS
* The initial evidence suggests that the number of bacterial species, a high number of bacteria, and the presence of anaerobes may be contributing factors to MFW symptom occurrence and severity.
* Developing a psychometric tool that comprehensively measures all symptoms of MFWs is imperative.
* Research using modern genomic technology is urgently needed to comprehensively examine the microbiome and metabolome of MFWs.
* Comprehensive examination of the microbiome and metabolome of MFWs and their association with symptom characteristics is the first step to identify targeted systemic and topical microbicides and therapeutics.
ACKNOWLEDGMENT
Thank you to the American Cancer Society (ACS): DSCN-18-214-01-SCN 250 Williams St, NW, 6th Floor, Atlanta, GA 30303.
REFERENCES