Sleep medication

The prescription of sleep medication in age group 15–29 was higher in HS population, where the prevalence lies at 8.1% (7.4% in men and 8.4% in women), compared to 3.0%

(2.6% in men and 3.5 % in women) in the general population. The prevalence then increases in age group 30–59 of both populations, with a prevalence of 16.8% (13.4% in men and 17.8% in women) in HS patients, and 7.2% (5.5% in men and 8.9% in women) in the general population. A large increase is also observed in this age group 60+ yr. in both HS patients and in general population. In HS patients of the age group showed a prevalence of 27.3% (18.7% in men and 31.1% in women), compared to 19.2% (14.0% in men and 23.8% in women) in the general population.

Table 1. Mental health status in HS patient vs. Swedish population, divided into different age-groups.

Psychiatric comorbidities are clearly more common in HS patients in all age-groups and in all diagnoses.

Fig 1. Mental health status in age group 15-29 in HS population compared to general population in Sweden. In this age-group like in other age-group we can see a significant increase in mental health disorders.

10 DISCUSSION

Study I

Hurley published his classic scoring system (33), and it does provide an good clinical overall staging of disease severity. However, it is lacking in detail and dynamics. Among patients with HS referred to a dermatological clinic, cases graded as Hurley II form the majority, and within this group there is a wide range of clinical findings and symptoms. Milder cases with accompanying smaller problems exist in this group, as do cases with very severe symptoms.

This variation is important to describe, especially in clinical studies, and therefore HSS was developed, a more detailed and dynamic scoring system for HS taking into consideration the inherently multifocal nature of the disease. By assigning numerical scores to affected regions as well as characteristic disease elements, it was proposed that the disease intensity could be quantified and integrated in a clinically more meaningful way. In addition, outcome measures should ideally have a high ICC to ensure that the findings are reliable. For the HSS the ICC was found to be 0.95 which is considered a near perfect score. In comparison, the Nail Psoriasis Severity Score, for example, has a reported ICC of 0.65–0.78 (73). The HSS is designed primarily for clinicians especially interested in HS, and the low interobserver variability suggests that the system is robust. Degrees of responsiveness to known risk factors and treatment are also important elements in the validity of a clinical measure. We have previously reported HSS values in 115 patients with HS, showing significant differences between the groups according to Hurley stage, between non-smokers and smokers as well as between normal weight women compared with obese women (68). There was also a positive correlation of fair degree between HSS and DLQI (68); however, quality of life scores in the literature are described to correlate poorly with objective disease assessment (74, 75).

Negative experiences with the HSS have, however, also been reported: Brunasso et al.(76) found the scoring system very difficult to apply in severe HS. There are several limitations in the use of HSS. Firstly, we have noticed that for the most severe Hurley II and III cases with widespread disease that affects large skin areas, the interobserver variability increases.

Secondly, the degree of inflammation of individual lesions is not included in the score, which would be valuable when applied to nonsurgical therapies or studies of risk factors. Thirdly, when completing the protocols, for some cases the observers had difficulties in distinguishing between a small fistula and a large nodule. This is important for the total score, as a fistula is scored as 6 points and a nodule only as 1. Technical aids e.g. ultrasound, which may be useful to distinguish the elements are, however, not practical in a routine setting (77), and more

stringent clinical definitions may therefore have to be developed in this field. Further, it is obvious that in the special setting of this study, with an initial training session, the variability obtained was low. In the routine clinical situation, without the training, it seems reasonable to believe that the variability will be somewhat larger. Finally, subjective variables are not included in the total score. Since pain is a major feature of HS it can be added by grading the worst lesion chosen by the patient using a visual analogue scale. DLQI or other subjective scoring systems may be used to complement the HSS in future prospective trials.

Further attempts to develop scoring systems more easily used in studies of HS have been made. The Hidradenitis Suppurativa Physician´s Global Assessment scale (HS-PGA)

evaluates number of nodules, abscesses and fistulae and divides severity into six levels (clear, minimal, mild, moderate, severe, and very severe) (35).

To meet the needs of measuring treatment effects in studies another method was developed, The Hidradenitis Suppurativa Clinical Response (HiSCR) (36) which measures clinical response and evaluates decrease in inflammatory nodules and abscesses. The score is however not suitable for example when measuring disease burden in a population HSS is simple to use and shows low interobserver variability. The score correlates with suggested risk factors, indicating that it reflects a valid estimation of disease severity.

The work on the optimal scoring system is still ongoing but HSS is one of the most reliable scoring systems.

Study II

A large variety of microorganisms can be isolated from HS lesions. Various bacteria are suspected of being responsible for parts of the inflammatory process (78). In many cases, the lesions have been found to be sterile (20, 79) but in some studies a large variety of

microorganisms have been isolated from the sinuses, particularly staphylococci, streptococci, Gram-negative rods and anaerobic bacteria(78).

The results in our study show that the bacterial cultures were positive for various

microorganisms and that all the patients had a polymicrobial flora with up to eight species.

High numbers of CNS were isolated in all patients, both from superficial and deep sampling.

CNS are commensals of the skin and mucosa and have long been dismissed as culture contaminants; however, they can cause severe infections in immune-suppressed patients, as

well as patients with intravascular devices and foreign polymer bodies (80). In recent years, several CNS species have been implicated in human infections. Most important in the pathogenesis of foreign-body-associated infections is the ability of these bacteria to colonize the polymer surface by the formation of a thick, multi-layered biofilm (80). In a former study performed at our lab, the bacteriology of chronic HS was addressed in 25 cases (20). To avoid contamination from the superficial bacteria in the overlaying skin, the same scanned CO2 laser technique was used (20, 81). Cultures from one superficial and two subsequent deep levels were taken. The most common finding was CNS in 21 superficial and 16 deep levels (20), a finding that was comparable with our present results.

An interesting difference from our present study was the frequent occurrence of

Staphylococcus aureus, found in 14 superficial and six deep cultures of chronic HS (20). In the present study we could not detect S. aureus in any culture, either at superficial or deep level. Previously it has been suggested that S. aureus may be involved early in the disease pathogenesis, causing anatomical changes in the hair follicles by inflammation and

necrosis(79). Contrary to our findings, we had expected the recognized pathogen, S. aureus, to be commonly found in acute exacerbations of HS. Explanation for these findings came newly from Barmatz and colleagues who come upon a big difference in bacteriologic profile in HS patients depending on previously antibiotic treatment (82). Further, non-standardized antibiotic treatment of HS in the community setting alters the microbiology of skin lesions toward gram-negative bacteria.

Since the publication of our study several other studies have been published on this subject.

The methodology stretches from traditional microbiological techniques for culturing bacteria in aerobic and anaerobic environment to more advances new methods like using PCR and DNA sequencing analysis(83).

These studies could identify a broad spectrum of bacteria but most common was CoNS (34,1%) and mixed anaerobic bacteria (23.3%). S. aureus also was also found to be considerable part of the overall isolated bacterial (83).

Although S. aureus is frequently presented in other skin diseases (e.g., atopic dermatitis and psoriasis vulgaris) (84, 85), but the clinical picture in HS differs significantly. Patients with folliculitis decalvans (FD) express a hypersensitivity towards S. aureus (86). In FD, the interplay between S. aureus and a deficient host immune response contributes to development of suppurative and cicatricial alopecia. Hypothetically a similar pathogenetic correlation may play a role in HS.

An interesting field in this area is microbiome in HS which is under ongoing investigation (87). The use of antimicrobial agents in HS is a well-established treatment, despite HS now being recognized as an autoinflammatory rather than infective disease (88). The microbiome has been implicated in the pathogenesis of HS, although the interaction between organisms and host remains to be fully elucidated.

As discussed in a recent review by Wark et al. (87) of the microbiome in HS, Coagulase-negative Staphylococcus, anaerobes such as Porphyromonas and Prevotella,

and Staphylococcus aureus species were commonly presented across the included cutaneous microbiome studies. Further, one small study on gut microbiome demonstrated an alteration in the gut microbiome composition compared to controls. Another study in this field found indifferences in the gut microbiome in patients with HS, compared to those with

inflammatory bowel disease (IBD) and HS, and IBD and/or psoriasis (87).

Given the fact that cutaneous dysbiosis develops in HS lesions (89), rise a question whether it affects the course of HS or if it has no causative relevance. Does the cutaneous inflammation alter microbiome, or the microbiome modify cutaneous inflammation?

This hypothesis would be easier to investigate in the setting of adequate and effective immunomodulating treatment such as adalimumab and other biological therapies which are currently in late-stage clinical trials.

More studies are needed on this field to identify the relationship between microorganisms and host, and this may reveal new pathways of disease pathogenesis and may help identify

potential future therapeutic targets.

Study III

To establish a general concept about burden of disease in terms of negative effects on patients’ quality of life, socioeconomic costs and need for treatment it is of importance to know the prevalence of a disease in different regions or countries. We performed a large registry-based cross-sectional study describing prevalence and social and comorbidity characteristics of 13,548 HS patients in a Swedish cohort defined by record linkage from registers covering the entire Swedish population. Lifestyle factors for, to our knowledge, the largest studied cohort of pregnant women with HS (n = 1,368) were also characterized in a subgroup analysis. In general, studies on the American population seem to show lower

Cosmatos et al. (9) showed a prevalence of 0.053%. A recent study by Garg et al (8). on a large patient population (approx. 48 million) also showed a low prevalence 0.1% but could also show a difference in prevalence between the population with Caucasian versus African American origin. In small Danish studies the prevalence has been shown to be as high as 4.1 and 4.0% (7, 10). But other population studies, one also from Denmark(4), showed an

estimated prevalence of 2.1%, and another from France showed a prevalence of 1% (11). The difference in prevalence is possibly dependent on the method used for collecting data, as discussed. Our results show a HS prevalence of 0.14% (0.22% for women and 0.07% for men) and an expected 3:1 female-male ratio. The analysis of social variables confirms and further highlights the standing of HS patients as a vulnerable group from a socioeconomic perspective. In comparison to the Swedish reference population, HS patients are more often unmarried and have a lower education and lower income. These facts, for example, lower degree of education and lower income, are closely related and should of course be considered possible confounders to each other and not only related to the disease itself. However, our findings are in line with previous reports showing lower socioeconomic status of HS patients (90, 91). The 3% prevalence of comorbidity in inflammatory bowel disease (441 cases in 13,548 patients) indicates a higher occurrence than in the Swedish population, even if the aggregated reference data did not allow for age- or sex-adjusted comparison. This finding is in line with inflammatory bowel disease being a frequently reported comorbid condition in patients with HS and with the suggested common mechanism of disease(57, 92, 93). The 8%

prevalence of type 2 diabetes reflects the previously reported comorbidity in metabolic syndrome (14, 46, 57).

The registry-based analysis of pregnant women with HS adds to the knowledge about the challenges and the potentially higher risk these women face during pregnancy. The group displays a high prevalence of lifestyle factors strongly associated with negative impact on health and increased pregnancy-related risks, including high prevalence of overweight, obesity, drinking habits, and smoking. As shown by Sartorius et al. (68) and Kromann et al.

(46), these factors are also associated with more severe disease and worse prognosis (68, 94).

The risk factors are also considerably more prevalent in women with HS than in the reference population of Swedish pregnant women.

The measured prevalence of HS in this study is lower than several other estimates of the European prevalence of HS (14, 57, 95). A possible confounding factor is that the National Patient Register does not cover care given in the primary care setting, that is, HS patients who only had a registered HS diagnosis in primary care during the 14- year study period were not

included. Another possible contributing factor is that HS is often underdiagnosed due to lack of knowledge about the disease among other physicians, as has previously been discussed in the review by Dufour et al. (14). Since this study is based on registered HS diagnoses, it does not include individuals with HS who have not consulted health care for their symptoms, or have been treated under another, perhaps incorrect, diagnosis by another speciality. Another limitation of a registry-based study is that we cannot control that all diagnoses are set correctly; however, that hypothesis that is probably limited by the fact that the diagnoses were performed by dermatologists. Furthermore, there seems to be a gap, in general, between registry-based studies and self-reported questionnaire-based studies (91, 96). This is a factor to consider when interpreting the results. As mentioned above, there might be a risk that patients with more subtle symptoms were not be included in our registry-based study, since they would obtain specialized care to a lesser extent. This limits our material to individuals with more severe disease, and this might also confound the results of socioeconomic factors.

Another limitation of the study is the availability of data from the National pregnancy register. All study variables related to lifestyle factors were not registered for all pregnant women with HS and there is a potential for recall bias associated with this type of data. Also, information about the number of included pregnancies was not available for the aggregated reference data used in the comparison of smoking habits. In all, we do not consider these factors to constitute bias that skew the results in a substantial way in either direction. That adjusted comparisons with reference population were not possible in the analysis of comorbid conditions constitutes another limitation of the study. However, this large study is based on registers covering the entire Swedish population. It therefore includes all patients with a registered HS diagnosis in specialty care in Sweden. The 14-year period used for defining the HS population also contributes to securing that all HS patients were included. It is likely that a patient diagnosed with HS and who is not in total remission has at least one registration of the diagnosis during the 14-year period. However, the fact that there is a general diagnosis delay of 7.2 years (2) for HS is another fact that we must take into consideration, and there is a possibility that not all patients with HS during the later years of the study were included.

Also, the Swedish pregnancy register provides detailed information on lifestyle factors for pregnant women with HS. Swedish pregnant women are closely monitored in the maternity care system enabling close comprehensive comparison between the women with HS and the reference population of Swedish pregnant women. In all, the generalizability of the results should be considered high in Sweden since the entire population was studied. The results should also be considered relevant in other settings with comparable conditions regarding health care delivery systems and risk factors in the population. In conclusion, this study is, to

our knowledge, the largest investigation characterizing patients with HS. Despite the limitation of no statistical evidence, this study further adds to the knowledge about patients with HS as a vulnerable group with social and lifestyle factors that may have negative impact on health. It also provides insights into the risk factors prevalent in the population of pregnant women with HS. Increased understanding about key characteristics of the HS population is important for improving preventive measures and care. The results may, for example, be used to stress the importance of prioritizing weight loss and smoking cessation interventions in maternity care and other settings, and to improve patient information and treatment. The method of using record linkage of several national registers to define and study the HS population shows the potential to create a more complete understanding of variables important for improving all aspects of care for this vulnerable and often under-recognized group. The method could be used to include more registers and to study other aspects of the HS population, including health care utilization, adherence to guidelines for pharmacological and other treatments, welfare benefit usage, and comorbidity

Study IV

Further, the large material from the HS population in Sweden showed that patients diagnosed with HS suffers more from overall psychiatric diagnoses than the general

population. Depression and anxiety disorders were overrepresented in the HS patient group compared to the general population. This finding is supported by the increased use of psychotropic drugs, also found in our study.

Interestingly, also personality disorders and neuropsychiatric disorders were increasingly found in the HS patient group. The early onset and the stigmatizing aspects of the disease could be an explanation for the high prevalence of such psychological disturbances.

Furthermore, the delay of diagnosis (2) and difficulty to get relevant medical care can also have significance. The age of disease onset is for many patients during puberty and the disease itself as well as diagnostic delay (2) in this sensitive age might also add to the risk of developing psychiatric comorbidities later in life.

The study shows the prevalence of overall psychiatric diagnoses increased in the age group 30–59, and then stabilized or decreased in the age group 60+yr, for both HS population and the general population. However, the prevalence of psychiatric diseases in the HS group is clearly higher. One explanation to the increase in this age group may be the combination of

having HS in addition to other stressors in this period of life (career, meeting life partner, building family, having children etc.) might result in an additive negative effect resulting in higher prevalence of psychiatric illness (97).

Previous studies show either conflicting results regarding depression and anxiety as comorbidities of HS (98), or unreliable results due to the small groups that were included in these studies. In contrast, our study included many HS patients (13538), and a reference population comprising the whole Swedish population (9 million). Each patient of the HS population is well documented by a unique personal number. Moreover, the populations studied had a wide age range (15-60+). These factors increase the reliability of our present results.

Compared to studies from the UK (98), we include all HS patients from the one and only national register and the control group was the whole Swedish population. This is also unique, compared to a recent finish study, which used patients with psoriasis and benign nevi as a control group (98).

In the meta-analysis of Machado et al (99) the prevalence of depression and anxiety was investigated as comorbidities in adults with HS. The studies included were either cross-sectional, retrospective, or prospective cohort studies with a setting of population based on in/outpatients. In our present study all patients that were diagnosed within the study period were included in the material, independent of the activity of the disease or whether the disease become inactive. Furthermore, we must address the bias of only having access to summary data for the Swedish population, and therefore having to suffice with using a description of the findings, as it was not possible to perform a complete statistical analysis.

Our present findings are highlighting the importance of early diagnosis of HS and raising awareness of its consequences on mental health. This is yet another challenge in the clinical daily practice. Furthermore, it emphasizes the importance of a multidisciplinary approach in clinical care.

11 CONCLUSIONS

HS is one of the most handicapping skin affections, resulting in multiple emergency hospitalizations, surgeries, sick leaves, and invalidity (57). HS disease is unfortunately not easily treated and there is a great need for research in this field.

In study I, the modified Hidradenitis Suppurativa Score (HSS) was evaluated and the interobserver reliability of the HSS described. Furthermore, its correlation with risk factors and disease severity was documented. The results suggest that HSS is simple to use and shows low interobserver variability, and it correlates with selected risk factors, indicating that it reflects a valid estimation of disease severity.

In study II, bacteria were isolated from acute nodules or abscesses during carbon dioxide laser vaporisation. Bacterial samples were taken from the skin surface (before surgery) and then from the deeper layers (during surgery). Coagulase-negative staphylococci were the most common bacteria, but contrary to what we expected, Staphylococcus aureus was not found in any cultures from acute inflammatory lesions of HS exacerbations.

In study III, a large registry-based cross-sectional study of social characteristics and comorbidity in all HS patients in Sweden as well as the prevalence of lifestyle factors in Swedish pregnant women with HS was performed. Results were compared with data from the normal Swedish population. The HS patients were more often women, unmarried, and had lower education as well as lower income. The pregnant women with HS showed higher prevalence of overweight, and smoking.

In study IV, the same large registry-based cross-sectional study as in study III was used to analyse the relation between mental health and the disease. The results were compared with data from the normal Swedish population. HS patients suffer more frequently from

psychiatric diseases, including depression, anxiety, mood disorders, autism, and ADHD. The findings are reflected by the fact that psychotropic drugs are more often prescribed to HS patients.

Many HS patients have a significant degree of morbidity as measured by DLQI. To evaluate this in a more objective manner HSS has been proposed, and it has low interobserver

variability and is easy to use. Involvement of bacteria in HS is under discussion, but CNS is the commonest bacteria found in deeper parts of HS lesions. Compared to the normal population, HS patients have lower income, lower education and are more often unmarried.

Overweight, obesity and smoking are more common among HS patients and so are psychiatric conditions, such as depression, anxiety, autism and ADHD. Dermatologists should be aware of the important association between HS and psychiatric disorders.

12 POINTS OF PERSPECTIVE

BACTERIOLOGY

A large variety of bacteria are found in HS lesions and most of them belong to the normal skin flora. In bacterial analysis of the skin surface (swab technique) there is a risk of contamination from the resident flora of the skin. Bacteria, especially CNS have been demonstrated by others and our group in deep parts of HS lesion (study II)(20), indicating their pathogenic relevance, although Staphylococci aureus are sometimes found during acute and short-lived exacerbations of HS, which would in those cases rather represent secondary infection, although secondary infection is rather seldom seen in HS. On the other hand, as in acne, it is possible that bacteria are involved in the primary inflammatory process of HS.

Antibiotics do not cure the disease, but antibiotic therapy may diminish symptoms, reduce pain, discharge, and malodour, by antibacterial or anti-inflammatory mechanisms. However antibiotic treatment, both local and systemic, has one major downside, especially when used in the long term, the risk of transfer of resistance to other bacteria and to other body locations.

In conclusion, there are still some missing puzzles about the role of bacteria in HS.

TREATMENT

It seems that definitive cure consists of surgery, and pharmaceuticals function either as complement or as preparation prior to surgery, although quit recently, in two large RCT, adalimumab, a TNF alpha inhibitor, had a significant effect in HS (100). Topical therapy is recommended as first-line of treatment, the actual choice of formulation vary from country to country, often depending on local tradition, availability, and legislature. In our tertiary

referral clinic, there is a long experience of azelaic acid use, as monotherapy or in

combination with topical clindamycin. The idea is to attack the early phase of HS, where the target is the follicular hyperkeratosis. In a small case series, we saw improvements (Emtestam et al, manuscript under preparation), especially in milder cases, and the method may be used to optimise patients prior to surgery. Also, over the years, we have noticed cases that have almost cleared after long term use of topical azelaic acid. These observations need to be studied in future properly designed RCT.