The role of bacterial colonisation in severity, symptoms and aetiology of hand eczema: The importance of Staphylococcus aureus and presence of commensal skin flora

The role and causality of the microbial ecosystem on the skin in relation to the development of hand eczema (HE) is still unknown.


| INTRODUCTION
Hand eczema (HE) is a common skin condition with a 1-year prevalence of 10% in the general population 1 and up to 21% in healthcare workers. 2 It is often characterised by a chronic relapsing course, as well as by significant psychological consequences for the patients. 3 The disease is of great socioeconomic importance 3,4 since it often leads to prolonged sick leave and sometimes even loss of job. 5 The pathogenetic aetiology of HE is complex and multifactorial. Impairment of the epidermal barrier function plays an essential role for the development of HE 6 and is affected by exogenous factors such as skin irritation and contact sensitisation 7,8 as well as by endogenous atopic predisposition. [9][10][11] In contrary to the well-established association between atopic dermatitis (AD) and Staphylococcus aureus (S. aureus) in regard to the pathogenesis 12,13 and the severity of the disease, 14 the bacterial colonisation of HE has only been sparsely investigated in the literature. A few previous HE studies found a high prevalence of S. aureus colonisation in HE patients, which was also associated with the severity of HE. [15][16][17] However, the role and causality of the microbial ecosystem, including S. aureus, on the skin in relation to the development of HE is still unknown.
The objective of this study was to investigate primarily the prevalence of different bacterial colonisations, especially with S. aureus, that are found in HE lesions in a large scale of patients and, secondly, their association with the severity, the clinical symptoms and the aetiology of HE.

| Patients
A total of 167 consecutive patients, who were referred to our department between September 2015 and December 2017 with the diagnosis HE were included in this retrospective, single-centre, cohort study.
The patient enrolment and the analysis of retrospectively collected patient data were approved by the institutional review board of our centre, and each patient consented to an anonymous data analysis.
This study was part of a project (Bactoderm), which collected skin swabs to establish a new gold standard method for bacterial PCRscreening. This new PCR-method did not meet its primary endpoint and therefore was not published.

| Patient data
Patient data from the first visit were included in the study. The data that were extracted from patients' files were demographic data, the HE subtype/aetiology, the symptoms of HE, the atopy status (AD, rhinitis allergica and/or asthma bronchiale) and the results of the microbiological investigation (isolation, cultivation and identification of bacteria). The atopic subtype included the cases with dyshidrotic atopic HE and with irritant-triggered atopic HE.
The clinical symptoms that were indicative of chronic HE were hyperkeratosis, rhagades and/or fissures and those indicative of acute HE were blisters, erosions and/or crusts.
The severity of the disease was assessed retrospectively by the treating physician (AB), based on the symptoms described in patient files: the cases with any of the symptoms of hyperkeratosis, rhagades, fissures, blisters, erosions or crusts affecting more than one finger were defined as severe, while all other HE cases were defined as mild.

| Statistical analysis
All statistical tests were performed by using SPSS (SPSS Inc). The Mann-Whitney U test was used to assess the association of S. aureus colonisation density with HE severity, primary HE aetiology (atopic HE vs all others), respiratory atopy in patients with non-atopic HE and clinical symptoms (acute, chronic or combination of the two), as well as the association of CSF density with HE severity. All tests were two-tailed, and p < 0.05 was considered statistically significant.
Furthermore, odds ratios (ORs) were calculated to investigate the influence of potential factors on S. aureus presence, HE severity and acute/chronic symptoms, respectively.  Table 1. 3.2 | S. aureus colonisation was associated with the severity of HE and acute HE symptoms Pathologic colonisation was significantly associated with the severity of HE ( p < 0.001; two-sided Fisher's exact test). The S. aureus colonisation, which represented the vast majority of the cases with pathologic colonisation (Table 1), was also found to be significantly associated with severe HE. Patients with severe HE had higher median S. aureus colonisation compared to patients with mild HE (Table 2). In addition, we found that acute symptoms of HE were significantly associated with S. aureus colonisation whereas chronic symptoms were not (Table 2).

| S. aureus colonisation was associated with atopic HE
Patients with atopic HE had a significantly increased risk of S. aureus colonisation and a higher median S. aureus colonisation compared to patients with other HE aetiologies (Table 2, Figure 1A). Respiratory atopy in patients with other HE subtypes/aetiologies did not increase the risk of S. aureus colonisation (Table 2).

| The presence of CSF had a protective effect on the severity of the disease
Patients with a positive test for CSF had a significantly reduced risk for severe HE (Table 2) and the absence of CSF was more common in severe HE cases than in mild HE cases (25.7% vs. 6.5%; Figure 2). Moreover, a higher concentration was associated with mild HE (Table 2). However, the majority of the patients with severe HE (74.3%) had positive tests for CSF (Figure 2). In addition, CSF and S. aureus in many cases occur together (Figure 3). We compared 35 patients with a high colonisation (++/+++) of both CSF and S. aureus with 36 patients who had a high colonisation of S. aureus but negative test for CSF concerning HE aetiology, severity and clinical symptoms and we found no differences (atopic HE in 97% vs. 97%, severe HE in 94% vs. 94%, acute symptoms in 71% vs. 75% and chronic symptoms in 37% vs. 34% of patients).

| DISCUSSION
The bacterial colonisation of HE has been only sparsely studied in the literature and those previous studies were mostly focused specifically on the S. aureus colonisation or on the atopic HE subtype. 17,20 Our study is the first to investigate the bacterial colonisation of the disease in a larger sample size of patients with various HE aetiologies (N = 167).
In accordance with previous studies, 15,16 we found a remarkably high prevalence of S. aureus skin colonisation in HE patients in comparison to the prevalence found in healthy individuals in the literature. 21 Colonisation with other bacteria species was found very sparsely. The presence of CSF was also frequent in the bacterial cultures, which was not described in other HE studies.
It should be highlighted that the present study investigated the bacterial colonisation and not the whole skin microbiome of the

| S. aureus colonisation and severity of HE
In agreement with previous studies, [15][16][17]20 we found that S. aureus colonisation was significantly associated with severe HE and that patients with severe HE had a higher density of S. aureus colonisation. S. aureus is an important pathogenic factor for HE and its effect on the severity of the disease may be due to superantigens from exotoxin-producing S. aureus strains which penetrate the skin barrier and contribute to the persistence and exacerbation of skin inflammation. 20 Alternatively, the causal link between HE severity and S. aureus colonisation may be the other way around, since the severity of HE and, thereby, the degree of an altered epidermal barrier function observed in HE, 6 might increase skin susceptibility towards S. aureus. 17 One study limitation that has to be noted is the high frequency of severe HE cases (81.4%), which could also explain the even higher prevalence of S. aureus colonisation (78.4%) in our cohort in comparison to previous studies (48%-69%). 15

| S. aureus colonisation and acute versus chronic HE symptoms
This is the first HE study to show that S. aureus colonisation is associated with acute symptoms (blisters, erosions and crusts) and not with chronic symptoms (hyperkeratosis, rhagades and fissures).
A similar result was found in a previous AD study in acute and chronic lesions. 22 In this study, the authors concluded that, in acute lesions of AD patients, there is a more direct damage to the epidermal skin barrier, mainly through scratching, and S. aureus can more easily penetrate or colonise the skin than in chronic lesions, which can also explain this association in HE patients. Another study showed a similar finding in patients with acute and chronic perianal eczema. 23 This finding may enable a personalised treatment of the patients depending on the clinical symptoms of the disease.

| S. aureus colonisation and atopic HE
As expected, considering the well-described association between S. aureus colonisation and AD, we found a significantly higher density of S. aureus in patients with atopic HE in comparison to other HE subtypes and we identified the atopic HE aetiology as a risk factor for S. aureus colonisation. This is in agreement with a previous study. 17 This correlation was found only in patients with atopic HE and not in patients with a history of respiratory atopy but in other HE subtypes/ aetiologies. However, while S. aureus colonisation was most common in atopic HE, it was frequently found in all subtypes of HE, which was also observed in the study of Nørreslet et al., 17

| Possible protective effect of CSF on HE severity
A new observation is also that the absence of CSF was more common in patients with severe HE than with mild HE, which may indicate a protective effect of CSF. Nevertheless, our study design cannot substantiate a causal relationship. Moreover, the presence of CSF excludes neither a severe HE case, nor the colonisation with S. aureus. The high ratio of patients who presented with severe HE despite an intact CSF may appear to contradict this protective effect. However, the majority of our patients had severe HE due to the setting at a tertiary referral centre and the ratio was significantly higher among those without CSF.
It was already shown that CSF protects from pathogens and plays a very important role in maintaining the balance of the immune system between effective protection and damaging inflammation. 25   as Cutibacterium and Corynebacterium, was even found in in-vitro and in animal studies to reduce S. aureus colonisation by forming porphyrin. 26 However, our data did not support this hypothesis since S. aureus and CSF often coexisted. In agreement with our findings, bacterial diversity was also closely linked in canine AD models 27 with the quality of the skin barrier, portrayed by transepidermal water loss and pH level, which also plays the most essential role in the pathogenesis of HE in all HE etiologies. 6 This protective effect of CSF could also be considered in future studies regarding new treatment options for HE.

| CONCLUSIONS
In conclusion, the main skin colonisation of patients with HE is with S. aureus and is associated with the severity of the disease, the atopic HE aetiology and the presence of acute HE symptoms. This is the first in-vivo study in humans to describe an association of CSF with a mild severity of HE. Future studies should further investigate this association and possible resulting novel management strategies.