Abstract
Background
Hidradenitis suppurativa (HS) is associated with inflammatory bowel disease (IBD), though risk factors remain to be determined.
Aim
To characterize HS among a cohort of IBD patients and identify risk factors for its development.
Methods
This was a retrospective case–control study at the ambulatory IBD centre at Mount Sinai Hospital from inception to May 2019. Patients with IBD who developed HS were included. Cases were matched 5:1 by age, gender (male versus female) and IBD type (ulcerative colitis [UC] or Crohn’s disease [CD]) to controls who had IBD without HS. Conditional logistic regression was used to calculate odds ratios (ORs) with 95% confidence intervals (95% CIs).
Results
Twenty-nine cases of HS (19 CD and 10 UC) and 145 controls were included. Of the 29 patients with HS, 11 (37.9%) were male and 18 (62.1%) were female. The severity of HS was mild in 10 (34.5%), moderate in 16 (55.2%) and severe in 3 (10.3%) patients. Patients with HS and IBD were more likely to be active (OR 10.3, 95% CI 2.0 to 54.0, P = 0.006) or past (OR 8.4, 95% CI 2.7 to 25.8, P < 0.005) smokers. Patients with HS and IBD were also more likely to have active endoscopic disease (OR 3.8, 95% CI 1.2 to 12.2, P = 0.022). Furthermore, those with HS and CD were more likely to have active perianal disease (OR 21.1, 95% CI 6.2 to 71.9, P < 0.005).
Conclusions
Active IBD, perianal disease and smoking may be associated with HS in IBD. Larger studies are needed to better characterize this morbid condition.
Keywords: Hidradenitis, IBD, Perianal, Skin rash
INTRODUCTION
Inflammatory bowel disease (IBD) is a chronic relapsing inflammatory condition that includes Crohn’s disease (CD) and ulcerative colitis (UC). Almost a third of patients with IBD develop extraintestinal manifestations (EIMs) (1), including musculoskeletal, ophthalmologic and hepatobiliary involvement (2,3). Furthermore, numerous mucocutaneous manifestations of IBD have been described such as erythema nodosum, pyoderma gangrenosum (PG) and aphthous stomatitis (3).
More recently, an association between IBD and hidradenitis suppurativa (HS) has been reported (4–7). HS is a chronic inflammatory follicular skin disease characterized by painful abscesses and sinus tracts in areas rich of apocrine glands such as the axillae, inguinal and anogenital regions. It is associated with significant morbidity and reduction in quality of life (8) and in rare cases, the development of squamous cell carcinoma (9). Common immunological, hereditary and microbiome pathways may predispose patients with IBD to the development of HS. In fact, the risk of HS among patients with IBD appears to be ninefold higher as compared to the risk in the general population, with a reported incidence of 0.55 cases per 1000 person-years (4,6,7). Similar to IBD, HS can also be associated with inflammatory conditions including spondyloarthropathies and is occasionally managed with biologic therapy (10). Active smoking (11) and obesity (4) appear to be strong risk factors for the development of HS in IBD. Furthermore, HS appears more commonly seen in patients with CD compared to UC, particularly in those with active fistulizing or perianal disease (11).
While the association between HS and IBD has been described, there is a need for more studies better characterizing the link between these two conditions. As such, we aimed to describe the clinical characteristics of the dermatologic features of HS in patients with IBD as well as IBD-specific features in these patients. Furthermore, we aimed to determine risk factors for the development of HS among patients with IBD, with a particular emphasis on disease activity.
METHODS
This was a retrospective case–control study performed at Mount Sinai Hospital, University of Toronto. It was approved by the research ethics board at Mount Sinai Hospital (MSH REB 18-0283-C).
Patient Selection and Data Collection
All ambulatory patients, aged >18 years of age, with UC, CD or indeterminate colitis followed in the clinics of the Mount Sinai Hospital IBD Centre, from inception to May 2019, were identified for chart review. Those with HS as diagnosed by the primary gastroenterologist or dermatologist on physical examination were included as cases in this study. In particular, patients with HS were identified by searching the electronic medical record for key terms such as ‘Hidradenitis Supprativa’, ‘Hidradenitis’, ‘suppurativa’, ‘skin condition’ or ‘skin boils’. These cases were matched 5:1 to controls with IBD who did not develop HS by age at IBD diagnosis, gender (male versus female) and IBD subtype (UC versus CD).
The electronic medical record was searched retrospectively for the collection of the following data: (a) baseline demographics including age, gender, most recent body weight or that at time of HS diagnosis, smoking status (no smoking history, past smoking history, active smoking at time of HS diagnosis), alcohol intake (reported by the patient on history and defined as any amount of alcohol intake at time of HS diagnosis) and other comorbidities (including respiratory diseases, such as chronic obstructive pulmonary disease, reactive airway disease or interstitial lung disease, cardiac diseases, such as ischemic disease or congestive heart failure, and any mental health disorder); (b) IBD characteristics including IBD subtype (UC versus CD versus indeterminate colitis), age at IBD diagnosis, current IBD-related medications at time of HS diagnosis, disease location, disease phenotype and disease activity (endoscopic, clinical or both) and (c) HS characteristics including timing of HS diagnosis (before IBD diagnosis, at time of IBD diagnosis, after an IBD diagnosis), location (axillae, inguinal, extremities, perianal or other), severity stage as defined by the Hurley system (stage 1: abscess formation without sinus tracts and scarring; stage 2: recurrent abscesses with sinus tracts; stage 3: diffuse interconnected sinus tracts and abscesses) (12) and HS-related treatment (antibiotics, local steroid injection, surgical excision or other modality). Finally, a dermatology consultation report, specifically addressing HS, was searched for in the medical record and additional data elements were collected if available.
Outcomes and Definitions
The primary outcome of this study was to better characterize the clinical characteristics of the dermatologic features of HS in patients with IBD as well as IBD-specific features in these patients. Risk factors for the development of HS in IBD were explored, with a particular emphasis on disease activity. Clinical disease activity was defined as a modified Harvey-Bradshaw Index (HBI) score >5 or partial Mayo (pMayo) score >2. For consistency, clinical disease activity was obtained either 3 months before or 3 months after the diagnosis of HS. Endoscopic disease activity was defined as the presence of any active inflammation (identified by the index endoscopist) on sigmoidoscopy or ileocolonoscopy within a calendar year before or after diagnosis of HS. For patients without HS, clinical and endoscopic disease activity was recorded in the same age year as the matched cases. Finally, the severity of HS was stratified by the aforementioned Hurley classification (12).
Statistical Analysis
Risk factors were defined as dichotomous values for statistical purposes. Categorical comparisons between cases and controls were analysed using the Pearson chi-squared test (if sample size was ≥10) or Fisher’s exact test (if sample size was <10). Median values with interquartile ranges (IQRs) were calculated for variables such as age, weight, disease activity scores and anti-tumour necrosis factor (TNF) drug levels. The Mann−Whitney U-test was used to calculate differences in these variables between patients who developed HS compared to matched controls. To determine specific associations with the development of HS in IBD, odds ratios (ORs) with 95% confidence intervals (95% CIs) were calculated using conditional binomial logistic regression analysis. A P value <0.05 was considered statistically significant. All analyses were performed using the SPSS software version 26 (Chicago, IL).
RESULTS
A total of 29 patients with IBD were diagnosed with HS and were matched to 145 patients with IBD without HS (Table 1). Of the 29 patients with IBD and HS, 11 (37.9%) were male and 18 (62.1%) were female. Ten (34.4%) were diagnosed with UC and 19 (65.6%) were diagnosed with CD. The median weight of patients with IBD and HS was 82.0 kg (IQR 41.5 kg) compared to 65.1 kg (IQR 21.5 kg) for those without HS (P = 0.008). Compared to those without HS, patients with IBD and HS were more likely to be current (4 of 29 [13.8%] versus 3 of 145 [2.1%]) or past (9 of 29 [31.0%] versus 9 of 145 [6.2%]) smokers (P < 0.005). Similarly, compared to those without HS, patients with IBD and HS were more likely to have any active alcohol intake (17 of 29 [60.0%] versus 43 of 145 [29.6%], P = 0.003) at time of HS diagnosis. There was no difference in the presence of other comorbidities between patients with IBD and HS compared to those without HS.
Table 1.
Baseline characteristics of patients with IBD at time of diagnosis with HS compared to those with IBD without HS
| Patients with HS, N = 29 | Patients without HS, N = 145 | P value | |
|---|---|---|---|
| IBD diagnosis, median age (IQR) | 20 (9) | 16 (8) | 0.098 |
| Gender | N/A | ||
| Male | 11 (37.9%) | 55 (37.9%) | |
| Female | 18 (62.1%) | 90 (62.1%) | |
| Smoking status | <0.005 | ||
| Current smoker | 4 (13.8%) | 3 (2.1%) | |
| Prior smoker | 9 (31.0%) | 9 (6.2%) | |
| No smoking history | 15 (51.7%) | 130 (89.6%) | |
| Unknown | 1 (3.5%) | 3 (2.1%) | |
| Alcohol intake history | 0.003 | ||
| Yes, current intake | 17 (60%) | 43 (29.6%) | |
| None or past history | 12 (40%) | 102 (70.4%) | |
| Weight (kg), median (IQR) | 82.0 (41.5) | 65.1 (21.5) | 0.008 |
| Other comorbidities | |||
| Cardiac | 2 (6.8%) | 3 (2.1%) | 0.194 |
| Respiratory | 2 (6.8%) | 3 (2.1%) | 0.194 |
| Mental health | 2 (6.8%) | 10 (6.9%) | 1.00 |
Values in bold are statistically significant. HS, Hidradenitis suppurativa; IBD, Inflammatory bowel disease; IQR, Interquartile range; N/A, not applicable.
Characteristics of Hidradenitis Suppurativa
All 29 patients with HS were diagnosed with their skin condition after being diagnosed with IBD. All patients were seen by a dermatologist for their HS. The median age of HS diagnosis was 25 years (IQR 10 years). Of the 10 patients with UC who developed HS, the areas affected were axillae only in 5 (50.0%) patients, inguinal only in 2 (20.0%) patients and both axillae and inguinal in 3 (30.0%) patients (Table 2). Four (40%) patients were treated with antibiotics for the HS, one (10.0%) had local steroid injections and four (40.0%) had surgical excision of the HS. The severity of HS was mild (stage 1) in four (40.0%) patients, moderate (stage 2) in five (50.0%) of patients and severe (stage 3) in one (10.0%) patient. Of the 19 patients with CD who developed HS, the areas affected were axillae only 2 (10.5%) patients, inguinal only in 3 (15.8%) patients, extremity only in 1 (5.3%) patient, perianal only in 2 (10.5%) patients and both axillae and inguinal in 8 (42.1%) patients. Ten (52.6%) patients were treated with antibiotics for the HS, two (10.5%) had local steroid injections and four (21.0%) had surgical excision of the HS. The severity of HS was mild (stage 1) in 6 (31.6%) patients, moderate (stage 2) in 11 (57.9%) of patients and severe (stage 3) in 2 (10.5%) patients. Overall, there were no differences in characteristics of HS in patients with UC and CD (Table 2).
Table 2.
Characteristics of HS in patients with inflammatory bowel disease
| Patients with HS | |||
|---|---|---|---|
| Ulcerative colitis (10 patients) | Crohn’s disease (19 patients) | P value | |
| Age at HS diagnosis, median (IQR) | 24.0 (10.0) | 25.0 (12.0) | |
| Location | 0.173 | ||
| Axillae only | 5 (50.0%) | 2 (10.5%) | |
| Inguinal only | 2 (20.0%) | 3 (15.8%) | |
| Extremities | 0 (0%) | 1 (5.3%) | |
| Perianal | 0 (0%) | 2 (10.5%) | |
| Axillae inguinal | 3 (30.0%) | 8 (42.1%) | |
| Other | 0 (0%) | 3 (15.8%) | |
| Treatment | |||
| Antibiotics | 4 (40.0%) | 10 (52.6%) | 0.518 |
| Local steroid | 1 (10.0%) | 2 (10.5%) | 0.965 |
| Surgical excision | 4 (40.0%) | 4 (21.0%) | 0.278 |
| Hurley stage | 0.562 | ||
| 1 (mild) | 4 (40.0%) | 6 (31.6%) | |
| 2 (moderate) | 5 (50.0%) | 11 (57.9%) | |
| 3 (severe) | 1 (10.0%) | 2 (10.5%) | |
HS, Hidradenitis suppurativa; IQR, Interquartile range.
Inflammatory Bowel Disease Characteristics in Patients With and Without Hidradenitis Suppurativa
The median age of IBD diagnosis in patients with HS was 20 years (IQR 9 years) and in those without HS was 16 years (IQR 8 years) (P = 0.098). Disease characteristics of CD and UC were compared between patients who developed HS and those who did not (Table 3). Compared to those with CD without HS, those with CD and HS were more likely to have a fistulizing CD phenotype (12 of 19 [63.2%] versus 21 of 94 (22.3%), P = 0.0003), history of perianal disease (16 of 19 [84.2%] versus 35 of 94 [37.2%], P < 0.0005) and active perianal disease at time of HS diagnosis (16 of 19 [84.2%] versus 15 of 94 [16.0%], P = 0.0005). There was no difference in disease location (ileal versus ileocolonic versus colonic) or prior CD-related surgery between the two groups. Similarly, there was also no difference in UC disease location or history of UC-related surgery in patients who developed HS compared to those who did not.
Table 3.
Characteristics of inflammatory bowel disease in patients with and without HS
| Patients with HS | Patients without HS | P value | |
|---|---|---|---|
| IBD patients | 29 patients | 145 patients | |
| IBD-related therapy | |||
| Corticosteroids | 4 (13.8%) | 10 (6.9%) | 0.256 |
| 5-ASA | 5 (17.2%) | 21 (14.4%) | 0.775 |
| Antibiotics | 3 (10.3%) | 1 (0.7%) | 0.015 |
| Methotrexate | 5 (17.2%) | 12 (8.3%) | 0.312 |
| Azathioprine | 1 (3.4%) | 13 (9.0%) | 0.470 |
| Anti-TNF | 16 (55.1%) | 72 (49.6%) | 0.840 |
| Infliximab | 12 (75.0%) | 52 (72.2%) | 0.525 |
| Adalimumab | 4 (25.0%) | 20 (27.8%) | 1.000 |
| Vedolizumab | 2 (6.8%) | 14 (9.7%) | 0.502 |
| Ustekinumab | 3 (10.3%) | 10 (6.9%) | 0.444 |
| Tofacitinib | 0 (0%) | 1 (0.7%) | 0.833 |
| Extraintestinal manifestations | |||
| Rheumatoid arthritis | 1 (3.4%) | 2 (1.4%) | 0.423 |
| Erythema nodosum | 3 (10.3%) | 5 (3.4%) | 0.132 |
| Other arthritis | 1 (3.4%) | 2 (1.4%) | 0.423 |
| Pyoderma gangrenosum | 2 (6.8%) | 1 (0.7%) | 0.072 |
| Psoriasis | 3 (10.3%) | 5 (3.4%) | 0.130 |
| Other | 1 (3.4%) | 16 (11.0%) | 0.312 |
| Clinical disease activitya | 14 (48.3%) | 35 (24.1%) | 0.008 |
| Partial Mayo score | 2 (IQR 5) | 0 (IQR 2) | 0.016 |
| Harvey-Bradshaw Index | 3.5 (IQR 8) | 1 (IQR 3) | 0.008 |
| Endoscopic disease activityb | 18 (62.0%) | 31/114 (27.2%) | <0.005 |
| Crohn’s disease | 19 patients | 94 patients | |
| Location | 0.135 | ||
| Ileal | 1 (5.3%) | 19 (20.2%) | |
| Ileocolonic | 15 (78.9%) | 52 (55.3%) | |
| Colonic | 3 (15.8%) | 23 (24.4%) | |
| Type | |||
| Stricturing | 5 (26.3%) | 30 (31.9%) | 0.788 |
| Fistulizing | 12 (63.2%) | 21 (22.3%) | 0.0003 |
| Hx of perianal | 16 (84.2%) | 35 (37.2%) | <0.0005 |
| Active perianal | 16 (84.2%) | 15 (16.0%) | 0.0005 |
| History of ICR | 2 (10.5%) | 23 (24.5%) | 0.236 |
| History of ileostomy | 0 (0%) | 7 (7.4%) | 0.351 |
| Ulcerative colitis | 10 patients | 51 patients | |
| Location | 0.159 | ||
| Extensive colitis | 5 (50.0%) | 37 (72.5%) | |
| Left-sided colitis | 3 (30.0%) | 11 (21.6%) | |
| Proctosigmoiditis | 2 (20.0%) | 3 (5.8%) | |
| History of IPAA | 0 (0%) | 5 (9.8%) | 0.580 |
Values in bold are statistically significant. 5-ASA, Aminosalicylate; HS, Hidradenitis suppurativa; IBD, Inflammatory bowel disease; ICR, Ileocecal resection; IPAA, Ileo-pouch anal anastomosis; TNF, Tumour necrosis factor.
aClinically active disease was defined as a Harvey-Bradshaw Index score >5 or partial Mayo score >2.
bEndoscopic disease activity defined by the presence of erythema, friability or ulceration at time of sigmoidoscopy or colonoscopy.
Compared to patients without HS, those with HS were more likely to be managed with antibiotics (3 of 29 [10.3%] versus 1 of 145 [0.7%], P = 0.015) for fistulizing complications (Table 3). There was no difference in anti-TNF use between the two groups (16 of 29 [55.1%] versus 76 of 145 [52.4%], P = 0.840). Furthermore, there was no difference in median anti-TNF drug levels at time of HS diagnosis compared to those without HS for infliximab (7.0 µg/g [10 patients] versus 7.9 µg/g [44 patients], P = 0.429) and adalimumab (12.0 µg/g [3 patients] versus 12.9 µg/g [14 patients], P = 0.950).
Finally, the presence of EIMs was compared between patients who developed HS and those who did not (Table 3). Compared to those who did not develop HS, patients with IBD and HS were numerically more likely to have a history of PG (2 of 29 [6.8%] versus 1 of 145 [0.7%], P = 0.072). Otherwise, there were no other differences in presence of EIMs in patients with IBD with HS compared to those without HS.
Disease Activity and Risk of Hidradenitis Suppurativa
Compared to those without HS, patients with HS were more likely to have clinically active IBD within 6 months of HS diagnosis (14 of 29 [48.3%] versus 35 of 145 [24.1%], P = 0.008) (Table 3). Similarly, compared to those without HS, patients with HS were also more likely to have endoscopically active disease (18 of 29 [62.1%] versus 31 of 114 [27.2%], P < 0.005). Specifically, in patients with UC, the median partial Mayo score in patients with HS was 2 (IQR 5) compared to a median partial Mayo score of 0 (IQR 2) in those without HS (P = 0.016). The endoscopic Mayo score in patients with UC and HS was 2 (IQR 1), compared to an endoscopic Mayo score of 0 (IQR 1) in those with UC but no HS (P = 0.031). For patients with CD, those who developed HS had a median HBI score of 3.5 (IQR 8), whereas those who did not develop HS had a median HBI score of 1 (IQR 3) (P = 0.008).
Logistic Regression Analysis
On conditional binomial logistic regression analysis, patients with HS and IBD were more likely to have a past history of smoking (OR 8.4, 95% CI 2.7 to 25.8, P < 0.005) or were actively smoking at time of HS diagnosis (OR 10.3, 95% CI 2.0 to 54.0, P = 0.006) (Table 4). Similarly, these patients were more likely to have active IBD (a composite of clinical scores and endoscopic assessment) (OR 5.4, 95% CI 2.3 to 12.5, P < 0.005). This held true for active disease as defined by endoscopy (OR 3.8, 95% CI 1.2 to 12.2, P = 0.022) but not by clinical scores (OR 1.2, 95% CI 0.4 to 3.9, P = 0.740). Furthermore, patients with CD and HS were more likely to active perianal disease (OR 21.1, 95% CI 6.2 to 71.9, P < 0.005) but not stricturing (OR 1.1, 95% CI 0.3 to 4.6, P = 0.899) or fistulizing (OR 0.4, 95% CI 0.1 to 1.5, P = 0.173) disease. There were no differences in IBD-related therapies or extraintestinal manifestations between cases and controls.
Table 4.
Logistic regression analysis for risk factors for the development of HS in IBD
| Odds ratioa (95% CI) | P value | |
|---|---|---|
| Smoking status | ||
| Never | 1.0 | N/A |
| Former | 8.4 (2.7–25.8) | <0.005 |
| Current | 10.3 (2.0–54.0) | 0.006 |
| Current alcohol use | 0.4 (0.2–1.2) | 0.102 |
| IBD activity | ||
| Active clinicalb or endoscopicc disease | 5.4 (2.3–12.5) | <0.005 |
| Active clinical diseaseb | 1.2 (0.4–3.9) | 0.740 |
| Active endoscopic diseasec | 3.8 (1.2–12.2) | 0.022 |
| Crohn’s disease | ||
| Stricturing disease | 1.1 (0.3–4.6) | 0.899 |
| Fistulizing disease | 0.4 (0.1–1.5) | 0.173 |
| History of perianal disease | 6.7 (2.1–21.8) | 0.001 |
| Active perianal disease | 21.1 (6.2–71.9) | <0.005 |
| History of ICR | 0.4 (0.1–1.7) | 0.197 |
| Ulcerative colitis | ||
| Pancolitis | 4.9 (0.6–37.1) | 0.121 |
| Left-sided | 2.4 (0.3–22.0) | 0.426 |
| IBD-related therapy | ||
| Corticosteroids | 0.5 (0.1–1.8) | 0.267 |
| 5-ASA | 0.6 (0.2–2.1) | 0.389 |
| Antibiotics | 0.04 (0.01–0.5) | 0.011 |
| Methotrexate | 0.3 (0.1–1.1) | 0.079 |
| Azathioprine | 1.6 (0.2–13.7) | 0.654 |
| Anti-TNF agents | 0.6 (0.2–1.5) | 0.250 |
| Vedolizumab | 0.8 (0.2–3.6) | 0.730 |
| Ustekinumab | 1.6 (0.4–6.3) | 0.507 |
| Extraintestinal manifestations | ||
| Rheumatoid arthritis | 0.9 (0.03–21.9) | 0.959 |
| Erythema nodosum | 0.4 (0.06–3.3) | 0.421 |
| Other arthritis | 0.6 (0.02–19.1) | 0.747 |
| Pyoderma gangrenosum | 0.1 (0.004–2.6) | 0.165 |
| Psoriasis | 0.3 (0.07–1.4) | 0.123 |
| Other | 6.8 (0.4–105.6) | 0.172 |
Values in bold are statistically significant. 5-ASA, Aminosalicylate; HS, Hidradenitis suppurativa; IBD, Inflammatory bowel disease; ICR, Ileocecal resection; IPAA, Ileo-pouch anal anastomosis; N/A, not applicable; TNF, Tumour necrosis factor.
aMatched for age, gender and type of IBD (ulcerative colitis vs. Crohn’s disease).
bClinically active disease was defined as a Harvey-Bradshaw Index score >5 or partial Mayo score >2.
cEndoscopic disease activity defined by the presence of erythema, friability or ulceration at time of sigmoidoscopy or colonoscopy.
DISCUSSION
The development of HS in IBD has been associated with significant morbidity. As such, determining specific risk factors for HS in this population may allow for early intervention which could lead to more effective treatment and improvement in overall quality of life. Previous studies have linked smoking and obesity as risk factors for the development of HS in patients with IBD (4,11). Here, we demonstrate that compared to those without HS, patients with HS and IBD are also more likely to have a smoking history, have active endoscopic IBD and have concurrent active perianal CD. Furthermore, there were no differences in overall HS characteristics in patients with UC and CD with respect to HS severity, location and overall management.
The association of HS in IBD appears to be stronger in patients with CD compared to those with UC. In fact, about two-third of the patients who developed HS in our study had CD, which is consistent with previous studies (7,13). Furthermore, fistulizing and perianal disease appear to be more common in patients with CD and HS compared to those who did not develop this skin condition. This may, however, be confounded by the fact that in routine practice, it may be quite challenging to distinguish perianal HS from fistulizing CD because of similar clinical presentation (11). In this setting, referral to a dermatologist and perianal imaging may assist in determining the correct diagnosis and appropriate management.
The association between CD and HS that has been demonstrated in the literature suggests common pathophysiological features between these two disease processes. Both CD and HS are characterized by the formation of sinus tracts and both share common risk factors such as smoking (14) and obesity (15). Moreover, common immunological pathways, such as dysregulation of interleukin-23/Th-17 and TNF-a, may be implicated in patients with both CD and HS (16). Previous studies have demonstrated that certain genes, such as SULT1B1 and SULT1E1 which are implicated in estrogen homeostasis, may be associated with the development of HS in CD (13). Finally, alterations in intestinal and skin microbiota may result in significant immune dysregulation, that is exacerbated by nicotine exposure, and an increased risk of HS in patients with IBD (17,18).
We also demonstrated that patients with IBD and HS were numerically more likely to have concurrent dermatological conditions such as PG, psoriasis and erythema nodosum compared to controls. Previous case reports and case series have suggested that patients with HS are at risk of developing PG (19,20). This also hints at common pathophysiologic underpinnings and a propensity towards developing dermatologic manifestations in a subset of IBD patients. As such, patients with IBD who develop HS or PG may benefit from close surveillance for the early detection of other potentially morbid skin conditions, particularly given their association with poor overall outcomes in IBD (21).
Consistent with previous studies, we also demonstrate that smoking and increased bodyweight may be associated with the development of HS in IBD (4, 11). In fact, Lukach et al. demonstrated that patients with HS and IBD were about 6 times more likely to have actively smoking at time of HS diagnosis and almost 11 times more likely to have concurrent diagnosis of obesity (11). Furthermore, compared to those who did not develop HS, we found that patients with IBD and HS were more likely to have active endoscopic disease. As such, close surveillance for dermatological diseases such as HS in patients with active IBD may allow for early identification and management of this morbid condition.
Interestingly however, despite this association of active disease with HS, we did not detect an increase in corticosteroid or biologic use in this cohort of patients. The literature is conflicting in the use of anti-TNF therapies in patients with IBD and HS. Whereas Janse et al. demonstrated that patients with HS and IBD were more likely to be prescribed anti-TNF therapy compared to those with IBD without HS (13), Lukach et al. did not find any significant difference between the two groups (11). These differences may be due to study design (questionnaire (13) versus case–control (11)) and patient characteristics. In our study, we included patients from a single tertiary care referral ambulatory clinic which may have enrolled patients with more aggressive disease phenotypes. This may have resulted in an increased use of anti-TNF therapy compared to the general gastroenterology practice in our study.
To our knowledge, our study is the first to correlate IBD activity with the development of HS. Given the relatively rare incidence of HS, we completed a retrospective case–control study to determine risk factors for the development of this morbid dermatological condition in IBD. Furthermore, all cases of HS were confirmed in consultation with a trained dermatologist at our university. However, there remains several inherent limitations. Our study was a single-centre study with a small number of patients diagnosed with HS. The small sample size might have resulted in type 2 error when assessing for associations between risk factors and development of HS in IBD. The retrospective nature of this study limited the completeness of available data and contributed to recall bias. For example, we were unable to determine whether the degree of severity of underlying IBD by endoscopic scores (i.e., simple endoscopic score for CD) and laboratory investigations (i.e., serum C-reactive protein or fecal calprotectin) is associated with an increased risk of HS. Similarly, variables such as the patient’s height were not recorded in all charts and as such, a body mass index could not be calculated. Other factors that may increase the risk of HS, such as oral contraceptive use and family history, were also not routinely available on chart review. Though we attempted to control for disease activity by recording clinical parameters at the same time point for all patients, because of the nature of this study, this reflects a static time point. Disease activity is a dynamic process and our analysis may not truly reflect the underlying inflammation at the time of HS diagnosis. Furthermore, given the case–control study design, we are unable to report any ‘cause-and-effect’ conclusions and can only identify possible associations that require confirmation in future prospective studies. Finally, we were unable to determine the effect of HS on the quality of life of IBD patients and whether therapeutic intervention improved the HS and associated morbidity.
Overall, we demonstrate that compared to those without HS, patients with IBD and HS are more likely to have active endoscopic disease, and in cases of CD, active perianal disease. Furthermore, patients with HS and IBD may also have other concurrent dermatological conditions such as PG though this did not reach statistical significance. As such, appropriate counselling and identification of specific disease features may allow for early detection of HS and implementation of appropriate therapy. More studies better characterizing HS in IBD patients are needed to further elucidate the shared pathophysiology between these two potentially morbid chronic inflammatory conditions.
Funding
None.
Conflicts of Interest: P.T., V.G. and Z.G. have no disclosure of potential conflicts of interest. A.W. has served on the Advisory Board of Janssen, AbbVie, Takeda and Ferring. A.W. has served as a speaker for Janssen and AbbVie.
Author Contributions: P.T.: Study concept design, data acquisition, data analysis, drafting of manuscript and critical revision of manuscript. V.G.: Data analysis, drafting of manuscript and critical revision of manuscript. Z.G.: Data analysis and critical revision of manuscript. A.W.: Study concept design, data acquisition, data analysis, drafting of manuscript and critical revision of manuscript. All authors have approved the final version of the article submitted including the authorship list.
References
- 1.Danzi JT. Extraintestinal manifestations of idiopathic inflammatory bowel disease. Arch Intern Med 1988;148(2):297–302. [PubMed] [Google Scholar]
- 2.Vavricka SR, Brun L, Ballabeni P, et al. Frequency and risk factors for extraintestinal manifestations in the Swiss inflammatory bowel disease cohort. Am J Gastroenterol 2011;106(1):110–9. [DOI] [PubMed] [Google Scholar]
- 3.Huang BL, Chandra S, Shih DQ. Skin manifestations of inflammatory bowel disease. Front Physiol 2012;3:13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Yadav S, Singh S, Edakkanambeth Varayil J, et al. Hidradenitis suppurativa in patients with inflammatory bowel disease: A population-based cohort study in Olmsted County, Minnesota. Clin Gastroenterol Hepatol 2016;14(1):65–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Kamal N, Cohen BL, Buche S, et al. Features of patients with Crohn’s disease and hidradenitis suppurativa. Clin Gastroenterol Hepatol 2016;14(1):71–9. [DOI] [PubMed] [Google Scholar]
- 6.van der Zee HH, de Winter K, van der Woude CJ, et al. The prevalence of hidradenitis suppurativa in 1093 patients with inflammatory bowel disease. Br J Dermatol 2014;171(3):673–5. [DOI] [PubMed] [Google Scholar]
- 7.van der Zee HH, van der Woude CJ, Florencia EF, et al. Hidradenitis suppurativa and inflammatory bowel disease: are they associated? Results of a pilot study. Br J Dermatol 2010;162(1):195–7. [DOI] [PubMed] [Google Scholar]
- 8.von der Werth JM, Jemec GB. Morbidity in patients with hidradenitis suppurativa. Br J Dermatol 2001;144(4):809–13. [DOI] [PubMed] [Google Scholar]
- 9.Fabbrocini G, Ruocco E, De Vita V, et al. Squamous cell carcinoma arising in long-standing hidradenitis suppurativa: An overlooked facet of the immunocompromised district. Clin Dermatol 2017;35(2):225–7. [DOI] [PubMed] [Google Scholar]
- 10.Miller IM, McAndrew RJ, Hamzavi I. Prevalence, risk factors, and comorbidities of hidradenitis suppurativa. Dermatol Clin 2016;34(1):7–16. [DOI] [PubMed] [Google Scholar]
- 11.Lukach AJ, Saul MI, Ferris LK, et al. Risk factors for hidradenitis suppurativa in patients with inflammatory bowel disease. Dig Dis Sci 2018;63(3):755–60. [DOI] [PubMed] [Google Scholar]
- 12.Goldburg SR, Strober BE, Payette MJ. Hidradenitis suppurativa: Epidemiology, clinical presentation, and pathogenesis. J Am Acad Dermatol 2020;82(5):. 1045–58. [DOI] [PubMed] [Google Scholar]
- 13.Janse IC, Koldijk MJ, Spekhorst LM, et al. Identification of clinical and genetic parameters associated with hidradenitis suppurativa in inflammatory bowel disease. Inflamm Bowel Dis 2016;22(1):106–13. [DOI] [PubMed] [Google Scholar]
- 14.van der Heide F, Dijkstra A, Weersma RK, et al. Effects of active and passive smoking on disease course of Crohn’s disease and ulcerative colitis. Inflamm Bowel Dis 2009;15(8):1199–207. [DOI] [PubMed] [Google Scholar]
- 15.Chan SS, Luben R, Olsen A, et al. Body mass index and the risk for Crohn’s disease and ulcerative colitis: Data from a European Prospective Cohort Study (The IBD in EPIC Study). Am J Gastroenterol 2013;108(4):575–82. [DOI] [PubMed] [Google Scholar]
- 16.Schlapbach C, Hänni T, Yawalkar N, et al. Expression of the IL-23/Th17 pathway in lesions of hidradenitis suppurativa. J Am Acad Dermatol 2011;65(4):790–8. [DOI] [PubMed] [Google Scholar]
- 17.Nikoopour E, Singh B. Reciprocity in microbiome and immune system interactions and its implications in disease and health. Inflamm Allergy Drug Targets 2014;13(2):94–104. [DOI] [PubMed] [Google Scholar]
- 18.Kurzen H, Kurokawa I, Jemec GB, et al. What causes hidradenitis suppurativa? Exp Dermatol 2008;17(5):455–6; discussion 457–72. [DOI] [PubMed] [Google Scholar]
- 19.Hsiao JL, Antaya RJ, Berger T, et al. Hidradenitis suppurativa and concomitant pyoderma gangrenosum: A case series and literature review. Arch Dermatol 2010;146(11):1265–70. [DOI] [PubMed] [Google Scholar]
- 20.Ah-Weng A, Langtry JA, Velangi S, et al. Pyoderma gangrenosum associated with hidradenitis suppurativa. Clin Exp Dermatol 2005;30(6):669–71. [DOI] [PubMed] [Google Scholar]
- 21.Weizman AV, Huang B, Targan S, et al. Pyoderma gangrenosum among patients with inflammatory bowel disease: a descriptive cohort study. J Cutan Med Surg 2015;19(2):125–31. [DOI] [PubMed] [Google Scholar]