Click here to load reader

Analysis of the chronic wound microbiota of 2,963 patients ... · PDF fileAnalysis of the chronic wound microbiota of 2,963 patients by 16S rDNA pyrosequencing ... does not improve

  • View
    217

  • Download
    0

Embed Size (px)

Text of Analysis of the chronic wound microbiota of 2,963 patients ... · PDF fileAnalysis of the...

  • Analysis of the chronic wound microbiota of 2,963 patientsby 16S rDNA pyrosequencing

    Randall D. Wolcott, MD1; John D. Hanson, PhD2; Eric J. Rees, PhD2; Lawrence D. Koenig, PhD2;Caleb D. Phillips, PhD2; Richard A. Wolcott, PhD2,3; Stephen B. Cox, PhD2; Jennifer S. White, MS3

    1. Southwest Regional Wound Care Center, Lubbock, Texas,

    2. Research and Testing Laboratory, Lubbock, Texas, and,

    3. PathoGenius Laboratory, Lubbock, Texas

    Reprint requests:Randall D. Wolcott, 2002 Oxford Ave,

    Lubbock, TX 79410., Tel: 806-793-8869;

    Fax: 806-793-0043;

    Email: [email protected]

    Manuscript received: September 16, 2015

    Accepted in final form: October 10, 2015

    DOI:10.1111/wrr.12370

    These authors contributed equally to this

    work.

    ABSTRACT

    The extent to which microorganisms impair wound healing is an ongoingcontroversy in the management of chronic wounds. Because the high diversityand extreme variability of the microbiota between individual chronic woundslead to inconsistent findings in small cohort studies, evaluation of a largenumber of chronic wounds using identical sequencing and bioinformaticsmethods is necessary for clinicians to be able to select appropriate empirictherapies. In this study, we utilized 16S rDNA pyrosequencing to analyze thecomposition of the bacterial communities present in samples obtained frompatients with chronic diabetic foot ulcers (N 5 910), venous leg ulcers (N 5 916),decubitus ulcers (N 5 767), and nonhealing surgical wounds (N 5 370). Thewound samples contained a high proportion of Staphylococcus and Pseudomonasspecies in 63 and 25% of all wounds, respectively; however, a high prevalenceof anaerobic bacteria and bacteria traditionally considered commensalistic wasalso observed. Our results suggest that neither patient demographics nor woundtype influenced the bacterial composition of the chronic wound microbiome.Collectively, these findings indicate that empiric antibiotic selection need not bebased on nor altered for wound type. Furthermore, the results provide a muchclearer understanding of chronic wound microbiota in general; clinical applicationof this new knowledge over time may help in its translation to improved woundhealing outcomes.

    The minor negative effects on wound healing resultingfrom the minimal colonization of certain microorganismsare supported by a significant body of literature. Severalhost-related factors can negatively affect wound healing,including microcirculatory impairment, endothelial celldysfunction, peripheral arterial disease, repetitive trauma,venous reflux, and poor nutrition. With regard to thisbroken host theory, it has been postulated that, once thebreech occurs, the impaired host environment allows forbacterial surface colonization that does not impair healing.This inconsequential presence of microbes is seen in caseswhere a specific bacterial species is cultured from achronic wound lacking any clear signs of infection, whentreatment of the patient to eliminate the identified microor-ganism(s) does not improve wound healing.1 Likewise, aCochran study concluded that there was no evidence sup-porting the routine use of systemic antibiotics to promotehealing in venous leg ulcers (VLUs),2 while a separatestudy determined that antibiotics should only be used totreat wound infections in diabetic patients, but not for sup-pression of bacterial colonization to promote wound heal-ing.3 Moreover, it has been proposed that treatment of thewound microbiome with antibiotics may comprise a con-tributing factor driving the observed increase in bacterialantibiotic resistance.4

    The SIDESTEP study highlights the confusing and oftencontradictory findings of randomized controlled trials uti-lizing cultivation methods.5 The authors of this studyfound that many MRSA-positive patients exhibited positiveresponses to antibiotic treatments that were insufficient forthis organism. Furthermore, this group demonstrated thatchronic wounds colonized by Pseudomonas spp. healedas well when treated with ertapenem, which has littleto no anti-pseudomonal activity, as those treated withpiperacillin/tazobactam (anti-pseudomonal therapy). Thisand other studies have led to the conclusion that certainbacteria, including pathogens such as Pseudomonas aerugi-nosa or enterococci, can colonize wounds without impair-ing wound healing.5 However, this position may fail tofully consider the polymicrobial nature of chronic wounds6

    as it is primarily based on the results of studies that haveutilized culture-based approaches that are inadequate forassessing polymicrobial samples. It is, therefore, possiblethat wound care management, when based on incompletediagnostics, may lead to suboptimal and confusing antimi-crobial outcomes.

    A second perspective is that the wound microbiota com-prises a major barrier to healing in any chronic wound.According to this viewpoint, chronic wounds are, in essence,chronic infections of the skin and adjacent tissues whose

    Wound Rep Reg (2016) 24 163174 VC 2015 by the Wound Healing Society 163

  • behaviors are in many instances directly related to the activ-ities of a polymicrobial biofilm.7 This view is predicated onthe fact that microorganisms (bacteria and fungi) use twodistinct infection strategies.8 Planktonic (free-floating)microorganisms are associated with classic acute infections,such as cellulitis, acute urinary tract infection, pneumonia,and sepsis, which are characterized by rapid onset and arobust host response (rubor, dolor, color, and tumor) that canoften be life-threatening. Typically, however, administrationof low minimal inhibitory concentrations of appropriate anti-biotics are required to eradicate the microorganism and,once cleared, the infection does not return. In contrast, theinflammation associated with chronic infections tends towax and wane. Moreover, while chronic infections oftenrequire very high doses of antibiotics for long durations (612 weeks), they typically respond incompletely to treatmentand reemerge once antibiotics are withdrawn. As such,infections are often clinically termed chronic once antibiotictherapy has failed.

    The difficulty in treating chronic infections is primarilydue to the ability of the infectious microorganism to pro-duce biofilms,7,9 which are polymicrobial communities(genetic diversity) in which each species exhibits quorumsensing control over gene expression (phenotypic diver-sity). Biofilm communities exhibit various characteristicsthat make them difficult to treat, including the slow pene-tration of antimicrobials, up-regulation of horizontal genetransfer in response to stress, anoxic cores, and the forma-tion of persister cells.10 Indeed, early studies showed thatantibiotics were only marginally effective against microor-ganisms within biofilms,11 and that biofilms are impervi-ous to both antibodies12 and white blood cells.13

    As the initial model of biofilm infection, the subcellularmechanisms by which bacteria attach to host tissues,14 uti-lize quorum sensing to control community-wide geneexpression,15 and induce inflammation to promote plasmaleakage from local capillaries for sustainable nutrition16

    have been elucidated. However, one of the most interestingmolecular strategies used by biofilm bacteria is the induc-tion of host cell senescence.

    There are various causes of wound bed cell senescencesuch as oxidative stress and host protease-mediated degra-dation of host cell receptors and/or cytokines. However, amore important and previously unknown cause for senes-cence occurs when biofilm bacteria use multiple smallmolecules to interfere with or commandeer the host cellprocesses, including rearrangement of the host cytoskele-ton,17,18 inhibition of mitosis,19 and, most importantly,inhibition of apoptosis.2022 Due to the wide array of path-ogenic effects exerted by distinct bacterial species, it maybe necessary to fully characterize the entire bacterial popu-lation of each biofilm.

    Biofilms often exhibit high levels of genetic diversityowing to the presence of multiple bacterial and/or fungalspecies, and this diversity provides numerous advantagesto the biofilm community. For example, diverse biofilm

    environments comprise large gene pools that allow formore efficient sharing of DNA sequences via horizontalgene transfer.23 Additionally, the microbial diversity of bio-films enables enhanced metabolic cooperation,24 byproductinfluences,25 passive resistance,26 and various other syner-gistic effects that provide the biofilm a competitive advant-age against the host.

    It is best to view biofilms as single entities that possessmultiple genetic resources, which allow them to adapt andeven thrive in the presence of various stresses. In general,increased genetic diversity imparts increased biofilm sur-vival.27 While individual biofilms almost always possess adominate microbial species, species that are present in lowabundance relative to the dominant organisms can have asignificant impact on the microbial community and caneven render the entire biofilm dysbiotic.28 Indeed, nor-mally nonpathogenic biofilms can cause disease in the hostdue to the activities of a minor constituent species. Thisfact adds great complexity to determination of the clinicalimportance of the microbes identified within woundbiofilms.

    Previous studies have shown that polymicrobial biofilmscan result in more severe infections that are more recalci-trant to treatment than monoclonal biofilms. Staphylococ-cus aureus biofilms containing low levels of P. aeruginosaexhibited increased rates of infection in a rat model,29

    while Prevotella increased the pathogenicity of S. aureusbiofilms in a mouse model of infection.30 Furthermore, P.aeruginosa waste products were shown to protect S. aureusfrom aminoglycoside-mediated killing.31 However, tounderstand whether the microbes