A Systematic Review and Critical Appraisal of Metagenomic and Culture Studies in Hidradenitis Suppurativa

Abstract

Background:

Hidradenitis Suppurativa (HS), also known as acne inversa, is a chronic inflammatory skin disease with still largely unknown pathogenesis. While infectious organisms have been identified in lesions of the disease since the 1980s, questions remain over the role that bacteria and microbiome play. Recent studies using 16S ribosomal RNA genesequencing and larger culture-based studies have begun to paint a clearer picture of the microbial world of HS. With this systematic review we summarizeall the work that has been done to date in HS bacteriology, analyze potential pitfalls and limitations of the current studies, and address future directions of investigation.

Methods:

This systematic review attempted to collate and analyze all bacteriology studies done to date. This review was prospectively registered with PROSPERO (1670769) performed in line with the PRISMA checklist.

Results:

22 studies were identified comprising 862 individual HS patients for culture studies and 206 HS patients for 16S rRNA genesequencing studies. Methodology tended to be varied, with different sampling, culturing and sequencing methods as well as amount of analysis and stratification of patients.

Conclusions:

Bacteria identified as elevated in HS lesions in sequencing studies as well as grown from HS lesions in culture studies are identified and discussed. These primarily included the anerobic Gram-negative bacilli Prevotella, Porphyromonas, and Fusibacterium, the Gram-positive bacilli Corynebacterium, and the Gram-positive cocci Staphylococcus, Streptococcus and Parvimonas. Potential interactions, as well as work in other disease models with related bacteria are also discussed. Areas of further investigation include in vitro studies of interactions between bacteria and keratinocytes, gut and oral microbiome studies and deep sequencing studies for virulence and phage factors.

INTRODUCTION

Hidradenitis Suppurativa (HS) is a chronic inflammatory skin disease with persistently draining purulent lesions of the axillary, gluteal and inframammary regions. New understanding of HS has begun to characterize the disease as a systemic auto-inflammatory phenomenon with possible environmental triggers, including a potential role for pathogenic bacteria.Proposed immunologic pathways include Th17/IL-23, TNF-a and the NLRP3 inflammasome¹. However, the role bacteria play in the pathogenesis of HS is still debated².HS tissue demonstrated dramatic elevations in antimicrobial peptides including S100A7–9, Lipocalin 2³ and beta defensins compared to normal tissue⁴, and the efficacy of antibiotics have been touted as evidence of a strong bacterial influence in disease activity². Are bacteria an infectious pathogen in HS, initiating or promoting the progression of the disease or does the development of sinus tracts provide a unique environment for new organism to flourish?

While no prospective, placebo controlled trials have been attempted with antibiotics, combined oral clindamycin and rifampicin improved up to 85% of patients in retrospective analysis of early HS lesions^5,6 and a combination of metronidazole, rifampin and moxifloxacin reached clinical remission in 75% of patients⁷. Considering the additional anti-inflammatory effects⁸ of the antimicrobial agents mentioned (regulation of macrophage function, NF-kb induction, reduced Th17 differentiation), the role of antimicrobial action on the microbiome specifically is still unclear. Recent culture-independent and culture-dependentstudies of the microbiomediscussed in this review provide additional information with some significant caveats in interpreting their results.

Most studies of the microbiome rely on targeted amplicon sequencing of the 16S ribosomal RNA gene hypervariable regions. However, choice of primers and variable regions can have significant impact upon the generalizability of results^9,10. One study of healthy cutaneous bacterial communities showed that V4 sequencing underrepresented Cutibacterium acnes and overrepresented Staphylococcus aureus. They concluded that V4 sequences strongly biased results in cutaneous samples¹¹ due to these differences in common skin microbiota. However, in a reply, Zeeuwen et al. suggested these shortcomings could be addressed with modification of the V4 primer pair¹². In addition, use of antibiotics prior to collection, emollients and hygiene all can potentially have effects on microbiome diversity results¹⁰. Other factors like PCR cycle length, DNA isolation, and amplicon preparation can also play significant roles.

For both microbiome and culture studies, the choice of collection can also influence results. Byrd et al. has recommended that for HS samples, skin swabs should be pre-soaked in enzymatic lysis buffer and that negative controls of ambient air be collected in addition to lesion and non-lesional sites¹³. However, a recently published study by Prast-Nielsen et al. found that microbiome populations differed significantly between swabs and punch biopsies in healthy skin (using V3-V4)¹⁴. Specifically, they found statistically significant differences between swabs and biopsies in the abundance of Bacteroidetesand Clostridiales as well as differences in α- and β-diversity. In addition, Schneider et al. found α-diversity differences in HS samples between cyanoacrylate biopsy and swabs¹⁵. This contrasts with an earlier and more limited study in 2008 that showed similar compositions between swabs, scrapes and biopsies in healthy controls¹⁶. Further validation will be needed to confirm these results, but this may influence future investigators in their choice of sample type when characterizing lesions. A brief summary of experimental design choices that can influence microbiome studies can be found in Table 1.

Table 1:

non-comprehensive list of patient and handling factors that can affect quality of data and evidence

Factors Effecting Microbiome Profiles	Common Variations
Patient/Disease Factors	Stratification and/orPrinciple Component Analysis (PCA) to identify differences between groups
Hurley Stage	Stage Stratifying from Stage I-III
Gender	Gender Stratification
BMI	Stratify categorically (>30) or as a continuous variable
Deodorant Use	Discourage use of deodorant before collection
Smoking	Record current/past smoking habits
Antibiotic use	All participants off active antibiotics for 1 week prior to collection
Processing Factors
Primer Choice and Design	V3-V464; V1-V365; V1-V266
Cycle Length	F3010
16S DNA yield/concentration	Various DNA extraction Kits
Sample type	Skin swab with buffer, punch or cyanoacrylate biopsy15

In contrast, whole metagenomic sequencing which uses shotgun unbiased sequencingof DNA has also been used to study the cutaneous microbiome in disease states like atopic dermatitis¹⁷. WGS allows for the opportunity to look at virulence factors and virusesbut comes with the disadvantages of cost and dealing with an overwhelming proportion of human DNA depending on the sample¹⁸. This can be partially abated with microbial enrichment of the samples but currently no published study has attempted shotgun metagenomic sequencing in HS lesional samples.

Aims:

Through this review of the literature we will attempt toanalyze the populations of commensal (those commonly found on skin) and potentially pathogenic bacteria (those previously linked to human disease) found in HS lesions as well as draw conclusions on the implications these organismsmay have on disease pathogenesis. In this review, we will highlight and critically evaluate recent work in next generation 16S rRNA sequencing as well as culturing of isolates. We will also discuss the potential shortcomings of different methodologies and identify knowledge gaps and areas for further research.

Methods

This systematic review was prospectively registered with PROSPERO (Registration Number 167079) and was conducted according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Figure 1). The PubMed and EMBASE database was searched for all primary literature in English language that addressed the overall bacteriology of HS from 1947 through December 2019. Terms used to find applicable studies included (((acne inversa) OR hidradenitis)) AND (microbiology OR bacteriology OR bacteria OR microbiota OR microbiome). Reviews, case reports, conference abstracts and those characterizing only a single or subset of bacteria were excluded. The reference sections of relevant articles were scanned manually to help find additional articles that may have not been capture via the initial search term. Data collection was performed independently by two authors and concordance of included papers was reached. Overall 22 papers were identified that matched our selection criteria and these were analyzed qualitatively due to differences in methods.

Figure 1:

PRISMA diagramdemonstrating the search strategy undertaken

RESULTS:

Metagenomic Sequencing Studies

Since 2017 there have been 5 published studies on the cutaneous microbiome of HS lesions, including 1 study that looked at the microbiome specifically in HS tunnels.Three of these studies sequenced the V3-V4 hypervariable region, one sequenced V1-V2 and oneV1-V3. Due to the differences in sequencing techniques, it is difficult to pool the results of these varied studies (Table 2).

Table 2:

16S Metagenomic Sequencing studies with identification of important bacterial differences and methodologies (as identified in Table 1)

Authors	Year	# of patients	16S Region	Sample Type	DNA Extraction	Analysis of Patient Factors?	Populations differing from controls
Ring 58	2017	30	V3-V4	Punch biopsy	DNeasy Mini Kit	Yes, Age, BMI, gender, HS, smoking	Porphyromonas↑, Peptoniphilus↑, Cutibacterium↓, Corynebacterium striatum↑
Guet-Revillet 66	2017	65	V1-V2	Swabs and punch biopsy	MagnaPure LC	Yes, correlations with BMI, duration of HS, Hurley Staging	Prevotella↑, Porphyromonas↑, Fusobacterium↑, Parvimonas↑, Corynebacterium↑, Staphylococcus↑
Naik 65	2019	11	V1-V3	Swabs	PureLink Genomic DNA kit	Yes: Hurley Staging	Prevotella↑, Porphyromonas↑, Fusobacterium↑, Clostridiales↑, Corynebacterium↑, Cutibacterium↓
Ring 67	2019	12 (Tunnels)	V3-V4	Swabs	DNeasy Kit	Yes, BMI, gender, smoking, age	Prevotella↑, Porphyromonas↑, Corynebacterium↑
Schneider 15	2019	11	V3-V4	Swabs, acrylate biopsies	DN easy Mini Kit	Yes, smoking, deodorant, age, gender, Hurley staging	Peptoniphilus↑, Porphyromonas↑, Corynebacterium↑ Finegoldia↑, Anaerococcus↑, Peptoniphilus↑ Cutibacterium↓,
Ring 68	2018	50 (serum)	V3-V4	Blood	DNA MiniKit	N/A	No differences found compared to controls
Hispan 69	2019	27 (serum)	16S	Blood	DNA blood minikit	Yes, age, BMI, smoking, Hurley staging	E. coli, Klebsiella pneumoniae, Gram-positive cocci.

Despite limitations in comparing and pooling data, the five NGS (Next Generation Sequencing) studies(n=129 total) done so far across many different geographical regions have elucidated common patterns in HS lesions. The most striking pattern to develop from all 5 studies is the greatly increased presence of Gram-negative and positive anaerobes, the increased presence of Gram-positive Corynebacterium and decreased presence of Cutibacterium(previously Propionibacterium)in active HS lesions. These findings in both Schneider and Ring’s studies may be partially attributable to bias in V4 primer sets as discussed previously. However,Naik et al⁶³ found similar results with different primer sets suggesting a more robust finding.

In addition, Naik et al and Guet-Revillet et al both found associations between anaerobic taxa and clinical severity of disease. Naik et al showed mixed anaerobic microbiota was associated with clinical severity and Guet-Revillet showed 11 taxa, most notably the anaerobes Fusobacterium and Parvimonas, to increase in abundance with higher classifications of Hurley staging.

Two of the studies reviewed (Ring 2017⁶⁵ and Schneider¹⁴) utilizedthe metagenome computational algorithm PICRUSt, (Phylogenetic Investigation of Communities by Reconstruction of Unobserved States)¹⁹. Thisalgorithm predicts the functional content of metagenomic 16S rRNA sequencing data using reference genomes and ortholog classification. The output estimates abundance of gene familiesrelated to metabolism, biosynthesis, cell cycle and DNA replication pathways. It is important to note that this does not directly look at cell metabolites and does not attempt to study the metabolomics of the underlying epidermis. Both studies identified quantitative changes in pathways related to cell growth, cell division, peptidoglycan synthesis and carbohydrate metabolism. These two studies, while limited, indicate a proliferative bacterial phenotype present in HS lesions.

One study by Ring et al. specifically investigated the microbiome found in the persistently draining sinus tracts, also known as tunnels, of moderate and severe HS patients. They found that these samples were heavily dominated by Prevotella, Porphyromonas, and other anaerobic bacteria. These anaerobes are not a major component of the microbiome of the healthycontrols in the other studies mentioned. In addition, in a separate non-sequence based study by Ring, tunnels had previously been shown to harbor bacterial biofilms using CLSM PNA-FISH.²⁰The biofilms were noted to be mostly negative for Staphylococcus specific probes and all were labeled as cocci suggesting that these biofilms might perhaps be composed of other cocci like Anaerococcus andParvimonaswhich have been found to have increased presence in some studies of HS lesions.

Two additional studies did 16S rRNA genesequencing of patient blood samples to find possible links between organisms present in serum and the cutaneous microbiome. The relatively small size of these studies and conflicting results prevents drawing conclusions from these studies, especially in light of the probability of contamination in blood draws.

Culture-Based Studies

Just as with sequencing studies of the microbiome, culture-based methods are also prone to bias from study design (Table 3). Because bacteria can grow at different rates with different nutritional requirements and tolerances, multiple growth conditions often need to be used in various culturing conditions (temperature and oxygen content, choice of media and various nutrient additions) in order to capture the entire spectrum of organisms present. For instance, without selective media for scarce organisms like Prevotella, more common bacteria or bacteria that grow more quickly can easily overwhelm a plate and prevent identification of rare species. Inherent to culture based studies, in contrast to NGS where all RNA present is amplified, is the fact that analysis is limited to organisms that grow and are viable and there are many reasons why an organism may or may not be present (sample quality, media choice, etc.), This limits direct quantification of communities. This, in conjunction with the diversity of methodology and lack of control group in most of the studies in table 3, limits analysis of cutaneous dysbiosis of HS. In the case of HS, where anaerobic Gram-positive and Gram-negative bacteria seem to have an increased presence and importance, anaerobic conditions with selective media for both Gram-positive and Gram-negative bacteria is essential to isolating these organisms.

Table 3:

Culture Studies to date with identification of methods and major bacteria identified as percent of isolates cultured.

First Author	Year	Patients	Sample Type	Analysis of Patient Factors	Culture Methods	Major Findings (% of isolates)
O’Loughlin 70	1988	27	Swab	No	unspecified	S. epidermidis (55%), S. pyogenes(19%), Proteus (11%), Bacteroides fragilis (3.7%)
Highet 71	1988	32	surgical swab	Yes, disease activity	Anaerobic, Aerobic	S. milleri (65%), S. aureus (43%), Bacteroides (31%), anaerobic streptococci (38%), Proteus mirabilis (28%)
Jemec 72	1996	41	Fine needle aspirate	Yes, collection site	Anaerobic, Aerobic(blood culture bottle)	CoNS (43%), S. aureus (28%), polymicrobial (14%)
Lapins 21	1999	25	Deep Swab	No	Anaerobic, Aerobic	CoNS (84%), S. aureus (56%), Peptostreptococcus (36%), Cutibacterium(28%), Prevotella (12%), Bacteroides (16%)
Brook and Frazier 73	1999	17	Needle aspiration/ swab	No	Anaerobic, Aerobic	Peptostreptococcus (59%), Prevotella (41%), S. aureus (35%), Bacteroides (24%), Streptococcus(17%)
Sartorius 22	2012	10	Biopsies from multiple levels	Yes, superficial/ deep	Aerobic, Anaerobic	CoNS (100%), Corynebacterium (90%), Anaerobic Gram-positive cocci (70%), Streptococcus (40%), Micrococcus (30%), Clostridium (30%), Prevotella (30%),Cutibacterium(30%), Bacteroides (20%),
Guet-Revillet 74	2014	82	Punch biopsies and swabs	Yes, disease severity	Aerobic, Anaerobic	S. lugdunensis (23%), mixed anerobic microbiota(74%), Anaerococcus (47%), Actinomycetes (87%), Peptoniphilus (42%), Prevotella (36%), Porphyromonas (27%), Bacteroides (25%), Streptococcus (77%),
Matusiak 75	2014	28	Swabs	Yes, disease duration	Aerobic, Anaerobic	S. epidermidis (22%), P. mirabilis (14%), S. aureus (14%), E. faecalis (12%), E. coli (5%), Streptococcus (10%)
Haskin 76	2016	189	unspecified	Yes, BMI >30	Unspecified	Corynebacterium (29%), Proteus (23%), S. aureus (23%), CoNS (22%), Streptococcus (21%), Enterococcus (14%), Prevotella (11%), Bacteroides (4%),
Hessam 77	2016	113	Deep swabs	Yes, gender, smoking, disease severity, antibiotic use	Aerobic, Anaerobic	CoNS (20%), S. aureus (13%), Proteus mirabilis (11%), E. coli (10%) Corynebacterium (6.4%), Enterococcus (6.4%), Streptococcus (12%), Prevotella (2.3%), Porphyromonas (1.8%), Bacteroides fragilis (1.8%)
Guet Revillet 66	2017	65	Swab, punch biopsies	Yes, disease severity	Aerobic, Anaerobic	Anaerobes (92.5%), Streptococcus anginosus (32%), Actinomyces spp (26%)., S. aureus (20.3%),
Nikolakis 78	2017	50	Swabs	Yes, disease severity, localization	Aerobic, anaerobic	Enterobacteriaceae (30%), Prevotella (28%), S. aureus (26%), CoNS (26%), Peptostreptococcus (14%), Streptococcus (10%), Bacteroides (10%),
Benzecry 79	2018	46	swabs	Yes, disease severity	Aerobic, Anaerobic	Streptococcus (26%), Corynebacterium (23%), Proteus mirabilis (22%), Peptostreptococcus (16%), GNR anaerobes (16%), S. aureus (16%)
Bettoli 80	2019	137	swabs	No	Aerobic, Anaerobic	Enterobacteriaceae (27%), CoNS (17%), Streptococcus (14%), Enterococcus (6.7%), Peptostreptococcus (4%), Bacteroides (3%), Prevotella (4%), S. aureus (6%), Peptoniphilus (4%), Pseudomonas (3%)

Between, 1988 and 2019 we found 14 relevant published studies that attempted to characterize the cutaneous microbiome of HS through culture. Earlier studies primarily identified various Staphylococcus and Streptococcus species but did not isolate any anaerobic Gram-negative organisms. This may in part be due to the evolution of microbiological techniques and agars. In addition, these Gram-positive cocci are predominant members of healthy skin microbiota and their presence is expected.

In concordance with the metagenomic studies, Gram-negative anaerobes were frequently isolated from patients with HS lesions. At least 11/14 studies isolatedPrevotella, Porphyromonas, and Bacteroides in significant quantities. Gram-positive anaerobic cocci of the genera Peptoniphilus, Anaerococcus and Peptostreptococcus were also very frequently found. In addition, although not identified in NGS studies, Enterobacteriaceae, especially Proteus mirabilis, were frequently identified in cultures. In addition, two studies by the same group (Lapins²¹ et al. and Sartorious²² et al.) investigated culture results in superficial and deep lesions. The earlier study by Lapins suggested that anaerobic bacteria were more often found in deep bacterial lesions while no significant difference was found in the second smaller study using CO₂ ablation to reach the deeper compartments.

DISCUSSION:

Discussion of Pathologic Relevance of Identified Organisms

Anaerobic Gram-Negative Rods

In terms of elucidating pathologic mechanism of disease in HS, the most promising group of organisms identified thus far in both culture-based and NGS studies are the Gram-negative anaerobes. This group of organisms is a common inhabitant of the oral and gut microbiome but as evidenced by the studies above is present in scant quantities as a member of healthy skin microbiota. They have been identified as a colonizer of HS lesions since at least 1980²³ and the organisms found comprise mostly of three genera: Porphyromonas, Prevotella, and Bacteroides. These three genera are all part of the family Bacteroidetes and were considered one genus until 1990²⁴ due to their similarities.

Prevotella and Porphyromonas are frequently associated pathogens in periodontitis²⁵, rheumatoid arthritis, metabolic syndrome and bacterial vaginosis²⁶ where they have been found to stimulate Th17 immune cells in vivo and in vitro.These pathogens drive Th17 resident memory cell expansion in vivo²⁷ dependent on IL-6 and IL-23 and promote Th17 differentiation in vitro via TLR2²⁸and has been shown to invade and promote the proliferation of gingival epithelial cells²⁹. Prevotella has also been shown to induce Th17 immune responses³⁰ and to successfully evade and kill neutrophilsin vitro³¹, which are an important part of HS immunity³².Both Prevotella and Porphyromonas species carry genes for excreted proteases which can act synergistically to cleave proteins like complement factor C3a and C5³³. However, while HS is increasingly shown to be IL17 driven, these organisms have only been observed to have increased presence in HS lesions observationally and no mechanistic studies have demonstrated a direct effect on lesions.

Aerobic Gram-Positive Cocci

Staphylococcus and Streptococcus are ubiquitous Gram-positive organisms on healthy human skin. Because of their common presence in samples, care must be taken to exclude its presence as due to contamination. Staphylococcus genus is traditionally divided into S. aureus, a common pathogen causing a range of infectious illnesses and a known haven of resistance genes (MRSA), and the commensal organisms grouped together as “coagulase negative” Staphylococcus (CoNS for their inability to produce an enzyme called coagulase). Both Staphylococcus aureus and Streptococcus pyogenes have been implicated in other autoimmune skin diseases like psoriasis vulgaris^34,35 and atopic dermatitis³⁶.In vivo studies showed that colonization with S. aureus induced a strong Th17 immune response³⁵.

Many of these Gram-positive cocci are rarely pathogens in otherwise healthy humans but have been repeatedly identified in skin and soft tissue abscesses and as opportunistic infections. Two studies not included in Table 2 specifically looked at the presence and virulence characteristic of S. lugdenesis³⁷ and S. epidermidis³⁸strains isolated from HS. These studies found increased resistance, growth and biofilm production in strains isolated from HS samples in comparison to strains isolated form healthy controls. The recent studies showing an increased virulence of S. lugdunensis and S. epidermidis hint at a potential role of these organisms, but further investigation is needed to more strongly prove a role in HS pathogenesis. Supporting a role of CoNS in HS is a revelation by Ring et al. that the sinus tracts that are present in severe HS infrequently contain CoNS biofilms²⁰.

In contrast to the NGS studies, many of the culture-based studies isolated Streptococcus and Staphylococcus in significant quantities which may be due to the biased limitations of culture vs metagenomics and ubiquitous presence of these organisms on the skin. However, this does not preclude their role in HS pathogenesis and the two studies of increased Staphylococcus virulence hint at this possibility^37,38.

Gram-Positive Rods

Another possible avenue that has been identified by the NGS studies is the possible interplay between the two Gram-positive rods, Cutibacterium(Propionibacterium) acnesand Corynebacterium striatum. Cutibacterium has been previously shown to be underrepresented by sequencing of V4 alone compared to V1-V3¹¹. However, a decrease in Cutibacterium and increase in Corynebacterium abundance in HShas been with a variety of primers which lends more confidence to the validity of the pattern. It is noteworthy that an increase of Corynebacterium in lesions is also seen in at least 3 studies of psoriasis vulgaris^39–42, which may reflect common microbial alterations across inflammatory skin diseases. The reasons for inconsistency between microbiome studies of psoriasis are beyond the scope of this review but highlight the need for understanding factors which contribute to variability in results across studies in HS.Cutibacterium is known as a major trigger for acne vulgaris and Corynebacterium striatum has been reported as anopportunistic pathogen. In addition, Cutibacterium creates propionic acid which lowers the pH of the surrounding microenvironment and has antimicrobial activity against other skin commensals like Staphylococcus aureus. Since Cutibacterium is decreased in HS lesions, it is possible that an increase in pH allows a cutaneous dysbiosis to develop and other pathogenic organisms to establish themselves in lesions^15,43.

Summary and Future Directions

In conclusion, through the multiple studies outlined in this review, a clear pattern of cutaneous dysbiosis develops in HS lesions which highlights the presence of anaerobic bacteria and a possibly virulent Staphylococcus species. Whether these observations are simply a bystander effect and have no causative relevance in HS remains to be elucidated. Is the microbiological dysbiosis causing the progression of HS, does it exacerbate and perpetuate the disease or is the dysbiosis simply a bystander effect of the newly formed sinus tract environments? This situation is similar to the ongoing debate regarding the pathogenesis of Atopic Dermatitis where a strong link to S. aureus has been established⁴⁴. It is unclear whether the microbiome influences cutaneous inflammation or if cutaneous inflammation alters the microbiome. Disentangling such hypotheses can begin by examining longitudinal microbiome studies in the setting of effective immunomodulating therapies such as monoclonal antibodies. This has provided great insights in AD^45,46 and can do the same in HS given that Adalimumab is currently licensed for the treatment of HS, and many other monoclonal antibodies are currently in late stage clinical trials^8,47–49.

One limitation of the current studies is a dearth of information on strain specific factors present in bacteria on HS skin lesions. Strain level tracking would require a combination of culture and whole genome sequencing. Bacteriophages and virulence factors can make significant differences in toxicity of an organism and even strain variation can change immunogenicity of two members of the same species⁵⁰. This information could help identify the specific drivers of HS inflammation.

Additional possibilities to explore associations with the HS cutaneous microbiome include in vitro assays with dead and alive bacteria and exploration of antibody responses to these putative pathogens. In addition to work mentioned above, work in periodontal disease has shown that gingival epithelial cells respond to presence of both Prevotella^51,52and Porphyromonas^53,54by increasing inflammatory cytokine expression as well as expression of metalloproteinases⁵⁵. These organisms can also neutralize innate immune mechanisms⁵⁶.Similar studies in both HS keratinocyte and normal keratinocyte samples should be undergone in order to support a pathogenic role of these organisms in disease.

None of the studies reported fungal dysbiosis, despite the important role that the Th17 immune response has in fungal immunity⁵⁷. Ring et al included 18S sequencing primers in their methodology to potentially capture fungal DNA but they did analysis only on Malassezia and did not report any significant results⁵⁸. It is likely that this lack of fungal data is due to fungal enrichment not being done prior to sequencing since the fungal mycobiome consists of less than 0.1% of the total microbial DNA load.Because of low burden of organism, any technique to characterize fungal changes in HS would need to specifically enrich for and target this population. However, Anti-Saccharomyces cerevisiae (ASCA) antibodies⁵⁹have been identified in affected patients. These ASCA antibodies have also been implicated in other auto-inflammatory disease like inflammatory bowel disease⁶⁰. In addition, a recent study was published that linked a yeast-exclusion diet to maintenance of remission after surgery⁶¹ providing a clue to a possible gut-skin axis in hidradenitis suppurativa.

In a similar vein, studies on gut microbiome and oral health which have not yet been done extensively in HS may provide important clinical clues to the sources of these bacteria. HS is strongly correlated withsmoking⁶², which have been linked with significant changes in the composition of the gut microbiome with increases of genera seen in HS lesions like Prevotella and Bacteroides.⁶³ Studies to compare strains found in the cutaneous and gut microbiomes of HS patients could potentially be lucrative in elucidating pathogenic mechanics.

CONCLUSION:

Thepathogenic significance of microbiome studies in HS is still unclear, withthe lack of standardization and heterogeneity of methods impeding overall generalization of results. It would be recommended for future studies to use a combination of culture based and NGS techniques (potentially sampling both swabs and biopsies when feasible), sampling from multiple disease sites in prospective longitudinal studies.

Additionally, in vitro keratinocyte experiments, strain identification, fungal isolation, gut microbiome sequencing, and long-term prospective tracking in the setting of active therapy are needed to clarify the role, if any, that bacteria have in the development and pathogenesis of HS.