been shown to significantly improve erythema (P=.003) and SCH (Central face, P=.041; Cheeks, P=.012).80 This may suggest that underlying (systemic) inflammation may exacerbate rosacea severity and EBD in a positive-feedback loop.
Microbiome Dysbiosis
Rosacea has several potential key microbes: Demodex spp. (Demodex folliculorum, Demodex brevis), Bacillus oleronius, Staphylococcus epidermidis, and Cutibacterium acnes.81
Demodex spp., specifically D. folliculorum and D. brevis, are commensal mites that dwell within the pilosebaceous follicle and consume sebum.82 The highest concentrations of Demodex spp. have been found in areas typically affected by rosacea (ie, cheeks, forehead).83 A 5-year epidemiological study of 3213 patients found PPR was associated with a mean Demodex density (Dd) of 36 D/cm2 (range: 8-112 D/ cm2) compared with ≤5 D/cm2 in normal skin.84 Non-invasive skin samples have found Demodex spp. more frequently in individuals with rosacea vs healthy age-matched controls (96% vs 74%, P<.01); in one study the Dd was ≥5.7x in rosacea patients.82 This appears to be unique to rosacea as other inflammatory conditions of the face failed to show excess Demodex presence.83 Clinicohistological studies have found mixed results but suggest that heavy Demodex spp. burdens may cause follicular rupture, dermal invasion, and cell-mediated reaction with granuloma formation along with activation of a plurality of inflammatory signals (including matrix metalloproteinase-9 (MMP-9), interleukin-8 (IL-8), tumor necrosis factor alpha (TNF-alpha), and LL-37).82,85 However, it is unclear if Demodex spp. instigates inflammation in rosacea or that underlying inflammatory changes enable proliferation of Demodex spp.86
Bacillus oleronius is a non-commensal gram-negative bacteria that has been cultivated from Demodex mite in PPR patients and is capable of stimulating a mononuclear cell response more readily in rosacea patients (n=16/22, 73%) vs healthy controls (n=5/17, 29%, P=.01).87 62 and 83 kilodalton (kDa) antigens isolated from B. oleronius have been found to preferentially stimulate serum reactivity in ETR individuals (n=21/26, 80%) over controls (n=9/22, 40%; P=.004) as well as in individuals with ocular involvement.88,89 Demodex spp. count was noted to be significantly greater in rosacea patients (6.6±9.0 vs 1.9±2.2, P=.014), suggesting Demodex may be a vector for B. oleronius induced inflammation.89 Inflammatory cascades downstream of B. oleronius include many factors associated with Demodex associated inflammation, including MMP-9, IL-8, TNF-alpha.90
S. epidermidis is a gram-positive coccus that, although capable of mitigating pathogenic S. aureus strain activity,91,92 has capacity for virulence.93,94 Cultures taken from PPR patients have found S. epidermidis within pustules but not from unaffected ipsilateral cheek skin.95 Furthermore, studies comparing S. epidermidis strains isolated from untreated PPR patients with controls found S. epidermidis from PPR pustules were significantly more likely to have β-hemolytic activity than from control isolates and also to secrete different (potentially virulent) proteins at higher temperatures (ie, 37℃ vs 30℃).96 Such temperatures are not uncommon in the vasodilated cheeks of a rosacea patients.
The lack of C. acnes may also play a role in rosacea.97,98 A small study (n=58) comparing superficial cutaneous swabs from individuals with PPR (n=15), ETR (n=21) with controls (n=22) found significantly less C. acnes isolated from the bilateral malar cheeks in rosacea patients (ETR: 27.3%; PPR: 23.3%) compared with controls (62.6%, P<.01).98 Another study found that healthy individuals age ≥60 have significantly less (P=.018) C. acnes than those <60.97 These observations may explain why AV is more prevalent among teens while another unrelated facial dermatosis, rosacea, is more common in those in the fourth/ fifth decade of life.
The authors also note that there is growing evidence of “crosstalk” between the enteric microbiome and CM in rosacea patients.99-105 Population studies have found individuals with rosacea were more likely to have celiac disease (Hazard Ratio 1.46, 95%CI 1.11-1.93, P<.001), Crohn’s disease (HR 1.45, 95%CI 1.19-1.77, P<.001), ulcerative colitis (HR1.19, 95%CI 1.02-1.39, P<.001), and irritable bowel syndrome (IBS) (HR 1.34, 95%CI 1.19- 1.50, P<.001).100 A randomized-control trial using rifaximin (a non-systemic, intestinal-limited antibiotic) for the management of small intestine bacterial overgrowth (SIBO) found that rosacea lesions completely cleared (n=20/28) or greatly improved (n=6/28) after treatment with rifaximin compared with those who received placebo (no change, n=18/20; worsened, n=2/20; P<.001).102 Other studies have also implicated alterations in the relative abundance of different species in rosacea including those of the genera Gordonia and Geobacilus, Corynebacterium, Actinomyces, Vellonella, and Chloroplast.103-105
Although these findings may imply a correlation between disease and dysbiosis, they cannot yet establish causality as it is unclear if dysbiosis is a primer or symptom of inflammation.104 Furthermore, studies have not consistently identified a single organism tying dysbiosis to rosacea, suggesting that dysbiosis itself may be sufficient to stimulate inflammation.103,104
Current Therapeutic Regimens and the Microbiome
Acne Vulgaris
Topical retinoids and benzoyl peroxide (BPO) are mainstays of mild-moderate AV management.106 BPO has been found to increase TEWL and deplete SC levels of α-tocopherol.107 Topical retinoids have been found to transiently thin the SC, increase cell turnover, and increase TEWL.18 Clinically, patients perceive these effects as irritation, inflammation, and xerosis, which may be partially alleviated by concurrent application of a gentle, non-comedogenic moisturizer.19-21 Topical treatment may also influence the microbiome (at least in the acute [post-]
Microbiome Dysbiosis
Rosacea has several potential key microbes: Demodex spp. (Demodex folliculorum, Demodex brevis), Bacillus oleronius, Staphylococcus epidermidis, and Cutibacterium acnes.81
Demodex spp., specifically D. folliculorum and D. brevis, are commensal mites that dwell within the pilosebaceous follicle and consume sebum.82 The highest concentrations of Demodex spp. have been found in areas typically affected by rosacea (ie, cheeks, forehead).83 A 5-year epidemiological study of 3213 patients found PPR was associated with a mean Demodex density (Dd) of 36 D/cm2 (range: 8-112 D/ cm2) compared with ≤5 D/cm2 in normal skin.84 Non-invasive skin samples have found Demodex spp. more frequently in individuals with rosacea vs healthy age-matched controls (96% vs 74%, P<.01); in one study the Dd was ≥5.7x in rosacea patients.82 This appears to be unique to rosacea as other inflammatory conditions of the face failed to show excess Demodex presence.83 Clinicohistological studies have found mixed results but suggest that heavy Demodex spp. burdens may cause follicular rupture, dermal invasion, and cell-mediated reaction with granuloma formation along with activation of a plurality of inflammatory signals (including matrix metalloproteinase-9 (MMP-9), interleukin-8 (IL-8), tumor necrosis factor alpha (TNF-alpha), and LL-37).82,85 However, it is unclear if Demodex spp. instigates inflammation in rosacea or that underlying inflammatory changes enable proliferation of Demodex spp.86
Bacillus oleronius is a non-commensal gram-negative bacteria that has been cultivated from Demodex mite in PPR patients and is capable of stimulating a mononuclear cell response more readily in rosacea patients (n=16/22, 73%) vs healthy controls (n=5/17, 29%, P=.01).87 62 and 83 kilodalton (kDa) antigens isolated from B. oleronius have been found to preferentially stimulate serum reactivity in ETR individuals (n=21/26, 80%) over controls (n=9/22, 40%; P=.004) as well as in individuals with ocular involvement.88,89 Demodex spp. count was noted to be significantly greater in rosacea patients (6.6±9.0 vs 1.9±2.2, P=.014), suggesting Demodex may be a vector for B. oleronius induced inflammation.89 Inflammatory cascades downstream of B. oleronius include many factors associated with Demodex associated inflammation, including MMP-9, IL-8, TNF-alpha.90
S. epidermidis is a gram-positive coccus that, although capable of mitigating pathogenic S. aureus strain activity,91,92 has capacity for virulence.93,94 Cultures taken from PPR patients have found S. epidermidis within pustules but not from unaffected ipsilateral cheek skin.95 Furthermore, studies comparing S. epidermidis strains isolated from untreated PPR patients with controls found S. epidermidis from PPR pustules were significantly more likely to have β-hemolytic activity than from control isolates and also to secrete different (potentially virulent) proteins at higher temperatures (ie, 37℃ vs 30℃).96 Such temperatures are not uncommon in the vasodilated cheeks of a rosacea patients.
The lack of C. acnes may also play a role in rosacea.97,98 A small study (n=58) comparing superficial cutaneous swabs from individuals with PPR (n=15), ETR (n=21) with controls (n=22) found significantly less C. acnes isolated from the bilateral malar cheeks in rosacea patients (ETR: 27.3%; PPR: 23.3%) compared with controls (62.6%, P<.01).98 Another study found that healthy individuals age ≥60 have significantly less (P=.018) C. acnes than those <60.97 These observations may explain why AV is more prevalent among teens while another unrelated facial dermatosis, rosacea, is more common in those in the fourth/ fifth decade of life.
The authors also note that there is growing evidence of “crosstalk” between the enteric microbiome and CM in rosacea patients.99-105 Population studies have found individuals with rosacea were more likely to have celiac disease (Hazard Ratio 1.46, 95%CI 1.11-1.93, P<.001), Crohn’s disease (HR 1.45, 95%CI 1.19-1.77, P<.001), ulcerative colitis (HR1.19, 95%CI 1.02-1.39, P<.001), and irritable bowel syndrome (IBS) (HR 1.34, 95%CI 1.19- 1.50, P<.001).100 A randomized-control trial using rifaximin (a non-systemic, intestinal-limited antibiotic) for the management of small intestine bacterial overgrowth (SIBO) found that rosacea lesions completely cleared (n=20/28) or greatly improved (n=6/28) after treatment with rifaximin compared with those who received placebo (no change, n=18/20; worsened, n=2/20; P<.001).102 Other studies have also implicated alterations in the relative abundance of different species in rosacea including those of the genera Gordonia and Geobacilus, Corynebacterium, Actinomyces, Vellonella, and Chloroplast.103-105
Although these findings may imply a correlation between disease and dysbiosis, they cannot yet establish causality as it is unclear if dysbiosis is a primer or symptom of inflammation.104 Furthermore, studies have not consistently identified a single organism tying dysbiosis to rosacea, suggesting that dysbiosis itself may be sufficient to stimulate inflammation.103,104
Current Therapeutic Regimens and the Microbiome
Acne Vulgaris
Topical retinoids and benzoyl peroxide (BPO) are mainstays of mild-moderate AV management.106 BPO has been found to increase TEWL and deplete SC levels of α-tocopherol.107 Topical retinoids have been found to transiently thin the SC, increase cell turnover, and increase TEWL.18 Clinically, patients perceive these effects as irritation, inflammation, and xerosis, which may be partially alleviated by concurrent application of a gentle, non-comedogenic moisturizer.19-21 Topical treatment may also influence the microbiome (at least in the acute [post-]
