Association Between Inflammatory Bowel Disease and Pruritus

Abstract Background Inflammatory bowel disease (IBD) is often complicated by extraintestinal manifestations. We frequently encounter IBD patients with pruritus; however, clinical evidence for the association of these conditions is lacking. Therefore, the present study investigated the incidence of pruritus in IBD patients. Methods Seventy-one IBD outpatients, including 55 with ulcerative colitis (UC) and 16 with Crohn disease, and 39 healthy volunteers (HVs) were surveyed about their pruritus symptoms using a visual analogue scale (VAS). Disease activities in UC and Crohn disease patients were classified according to partial Mayo and IOIBD (International Organization for the Study of inflammatory Bowel Disease) scores, respectively. Skin barrier condition was examined by measuring transepidermal water loss and stratum corneum hydration. The distribution of intraepidermal nerve fibers in skin samples from 9 UC patients was examined immunohistochemically using an antiprotein gene product (PGP) 9.5 antibody. Results Visual analogue scale scores were higher in IBD patients than in HV (P < 0.001). Active stage IBD patients had more severe pruritus VAS scores than those in the remission stage (P = 0.036). Transepidermal water loss was higher in IBD patients (P < 0.001) and active stage IBD patients (P = 0.004), while stratum corneum hydration was lower in IBD patients (P = 0.019) and active stage IBD patients than in HV (P = 0.019). A relationship was observed between the degree of pruritus and number of PGP9.5-immunoreactive intraepidermal nerve fibers in UC patients. Conclusions Inflammatory bowel disease patients, particularly active stage patients, frequently exhibit symptoms of pruritus and dry skin. This result may have predictive and therapeutic implications for the treatment of IBD symptoms.


Introduction
Inflammatory Bowel Disease (IBD) is often complicated by extraintestinal manifestations, including skin disorders, such as pyoderma gangrenosum, erythema nodosum, and erythema exudativum multiforme. 1 Furthermore, although anti-tumor necrosis factor (TNF)- biologics are commonly used in the treatment of IBD and psoriasis, paradoxical adverse skin manifestations have been reported during biological treatments for IBD. 2 These findings imply that some common immunoreactions are involved in IBD and its extraintestinal manifestations, particularly skin disorders.
Recent studies have suggested several causes of IBD, such as genetic factors (e.g. familial aggregation including twins and susceptibility genes), environmental factors (e.g. industrialization and a diet high in fat and sugar), and disturbances in the enterobacterial flora inducing intestinal barrier loss. 3,4 These causes, especially barrier dysfunction are considered to play multiple important roles in the progression of IBD.
Clinically, We frequently encounter IBD patients with pruritus and skin dryness; however, the most recent European Guidelines on Chronic Pruritus (2012) and other guidelines do not describe IBD-related itch. 5 This is an important issue because pruritus Downloaded from https://academic.oup.com/crohnscolitis360/advance-article-abstract/doi/10.1093/crocol/otaa012/5766370 by guest on 12 March 2020 A c c e p t e d M a n u s c r i p t 8

Evaluation of IBD, pruritus, and dry skin
IBD: Disease activity was classified according to the non-invasive 9-point partial Mayo score for patients with UC and the IOIBD score for those with CD. 7,8 The partial Mayo Scoring Index requires the patient to score two items, stool frequency and rectal bleeding, on a scale from 0 to 3, and for the physician to similarly score their global assessment. UC severity was evaluated as the sum of the three scores (maximum score of 9 points). The IOIBD score included 10 items each scored as zero for absence and 1 Downloaded from https://academic.oup.com/crohnscolitis360/advance-article-abstract/doi/10.1093/crocol/otaa012/5766370 by guest on 12 March 2020 M a n u s c r i p t 9 for presence including i) abdominal pain, ii) bowel evacuation six times or more per day, iii) anal lesion, iv) fistula, v) complications, vi) abdominal tumor, vii) body weight decrease, viii) fever of 38°C or higher, ix) abdominal tenderness, and x) hemoglobin level of 10 g/dl or lower. The sum of the 10 items was used as the IOIBD score for patients with CD.
We classified UC patients using the total point score of the partial Mayo score: 0-1, remission; 2-4, mild disease; 5-6, moderate disease; and 7-9, severe disease. In CD patients, an IOIBD score of 0-1 was deemed 'in remission', otherwise the disease was deemed 'active'. In the present study, since we were unable to perform endoscopy on every participant, disease activity was evaluated by clinical index only. The mean of the total partial Mayo score among 55 UC patients was 1.93 ± 0.33 (mean ± SEM), while that of the IOIBD index among 16 CD patients was 0.63 ± 0.22 (mean ± SEM).

Pruritus:
We asked participants to self-assess their severity of pruritus. The worst case of pruritus anywhere on the whole body in the preceding 3 days was evaluated using a visual analogue scale (VAS). Participants were asked to rate their pruritus state using a 100-mm VAS graded from 0 (no pruritus) to 100 (worst imaginable pruritus). The Dry skin: The skin barrier condition of all participants was evaluated by non-invasive measures of transepidermal water loss (TEWL) and stratum corneum (SC) hydration of the radial aspect skin of the right forearm. 10 Measurements were taken using VapoMeter ® and MoistureMeter SC Compact ® (Delfin Technologies Ltd., Kuopio, Finland), respectively. Those with presence of rashes, scratches, or other skin abnormalities in the measuring area or those that had applied any emollients in the measurement area were excluded. Since TEWL and SC hydration are often affected by temperature and humidity, 11 we took measurements in an air-conditioned outpatient clinical room during the period from winter to early spring (between November and March) when it was relatively cool and dry in Japan. The average air temperature in the clinic (measurement room) was 26.5 ± 0.14°C (mean ± SEM) and relative humidity was 25.7 ± 1.45% (mean ± SEM). We analyzed eosinophil counts and serum IgE to assess whether histamine was involved in induction of pruritus in IBD patients. Eosinophil counts were obtained using the ADVIA ® 2120i System (Siemens Healthineers AG, Munich, Germany) using peroxidase activity and nuclear density analysis cytograms. Blood samples were centrifuged at 3000 rpm at 20 to 25℃ for 10 min. Data on serum IgE levels were analyzed using an electrochemiluminescence immunoassay with a Cobas ® 6000/e601 analyzer (F. Hoffmann-La Roche Ltd., Basel, Switzerland). The normal range of serum IgE was defined as 0 to 232 IU/ml.

Immunohistochemistry
Three-mm punch biopsies were taken from the abdominal skin of nine UC patients (males 5, females 4, mean age 42.67 ± SEM 2.74 years, range 30 to 53 years). They had no clinical skin abnormalities such as rashes or infection. The protocol for immunohistochemistry of epidermal nerve fibers was described previously. 12 Briefly, to quantify the number of epidermal nerve fibers, five to eight specimens per skin biopsy were stained with the anti-product gene protein 9.5 (PGP9.5, 1:600; ENZO Life

Statistical analysis
Pruritus VAS scores, TEWL, SC hydration, IBD disease activity, and other results were analyzed using GraphPad Prism7 ® (GraphPad Software, La Jolla, CA, USA). Between group differences were analyzed using two-tailed Mann-Whitney test with P<0.05 deemed significant, and Spearman's correlation test was used to assess correlations among data with coefficients (r) 0.2 to 0.4, > 0.4 to 0.7, and > 0.7 deemed weak, moderate, and strong, respectively.

Ethical considerations
The present study was conducted according to the principles of the Declaration of Helsinki and was approved by the Medical Ethical Committee of Juntendo University Urayasu Hospital. All study participants provided written informed consent.

Results
The mean VAS score was significantly higher in IBD (mean VAS=32.74 ± SEM 3.58,  Figure 3D). Moreover, TEWL and SC hydration in the active and remission phases were not significantly different ( Figure 3C and 3D). We also classified TEWL and SC hydration by total partial Mayo scores in 55 UC patients ( Figure 3E and 3F). Remission (P=0.004), mild (P=0.009), and severe disease groups (P=0.020) showed significantly higher TEWL than HV. Only Severe disease patients showed less SC hydration than HV (P=0.020). No significant difference was observed in SC hydration between the other activity category and HV ( Figure 3F). However, a significant difference was noted between SC hydration in the moderate and severe disease groups (P=0.030) ( Figure 3F).   (Table 1 and Supplementary Table 1 (v)).

(See Supplementary
The medication of IBD is shown in Table1. None of the IBD patients had any specific extraintestinal skin manifestations. The mean duration of disease following diagnosis of IBD was 12.16 ± 0.84 (mean ± SEM) years(Table1).
The distribution of IENF in the skin of nine UC patients with pruritus was examined immunohistochemically using the anti-PGP9.5 antibody (Figure 4). PGP9.5-immunoreactive fibers in the epidermis and intraepidermal area exhibited higher densities in UC patients with pruritus ( Figure 4A-C) than in those without pruritus ( Figure 4D). The VAS score was associated with the number of PGP9.5-immunoreactive fibers in the epidermis ( Figure 4F) and intraepidermal area ( Figure 4G). No relationships were observed among the number of IENF, TEWL, SC hydration, or total partial Mayo score (Supplementary Table 3).

Discussion
Few studies have attempted to evaluate pruritus in IBD patients. The present study showed that the mean VAS score was significantly higher in IBD patients than in HV, and 68 % of IBD patients had mild to very severe pruritus (Figure 1). Furthermore, TEWL increased, while SC hydration decreased in IBD patients ( Figure 3A and 3B). This result suggests that IBD patients had dry skin due to barrier dysfunction. Patients in the present study were recruited from an outpatient clinic, and their IBD clinical conditions were generally mild (See : Table 1, mean of the total partial Mayo score and IOIBD score). We speculate that their dry skin condition was not related to dehydration due to frequent diarrhea. Dry skin induces pruritus. Therefore, these results suggest that most IBD patients manifest dry skin with pruritus, 13,14 . Pruritus worsens the quality of life of patients in addition to the discomfort associated with IBD 15, 16 , therefore pruritus needs to be considered in the management of IBD patients.
Previous studies demonstrated that dry skin induces itching, which is resistant to conventional treatments, such as histamine H1 receptor antagonists. 17 Furthermore, increased sensory nerve density in the epidermis is partly involved in itching A c c e p t e d M a n u s c r i p t 18 hypersensitivity, such as xerosis and atopic dermatitis (AD). 17 In the present study, we found increased IENF in UC patients, which was similar to dry skin features in AD. In a recent study, although the patient not noted as having a typical dermatological complication of IBD the study 18 reported a bidirectional association between AD and IBD. This finding raises the possibility that there is a close relationship between skin barrier and intestinal conditions. Moreover, our immunohistochemical findings showed a relationship between the degree of pruritus and number of IENF in UC patients (Figure 4). In contrast, no significant differences were observed in serum IgE levels or the eosinophil count between HV and IBD patients (Supplementary Table 2 (i), (ii)). These results suggest that many IBD patients frequently have dry skin and pruritus without marked dermatological manifestations and that itch in IBD patients is histamine independent.
The present results revealed that the disease activities of IBD correlated with the degree of pruritus, TEWL, and SC hydration (Figure 2A and 3C, D). Moreover, the degree of pruritus and TEWL were significantly higher in IBD patients than in HV, even in the remission phase (Figure 2A and 3C). These results imply that IBD patients We found the highest mean VAS score was in the moderate disease UC group rather than the severe disease group (Figure 2A Table 3). Therefore, although the number of skin biopsy samples examined was small, UC-associated pruritus may be an indicator of the worsening of symptoms in patients. A recent study on CD reported that extraintestinal manifestations, including skin disorders, occurred at a lower frequency in elderly-onset than in pediatric-onset patients. 19 Moreover, increased TEWL in IBD patients indicates skin barrier dysfunction, and an allergic reaction due to percutaneous sensitization for certain allergens may occur. 20,21 If IBD patients have skin barrier dysfunction, it may be associated with an immunoreaction due to percutaneous sensitization in addition to disruption of intestinal mucosal barrier and One of the limitations of the present study was that we were unable to perform colonoscopy on patients at the same time as assessing their skin condition. Thus, we did not investigate the relationship between mucosal and skin barrier dysfunctions. In addition, the serum analysis was limited by national health insurance coverage. In this study, the combination of IBD patient's medication was diverse even in remission, therefore the number in each group was too small to consider the effect of medications.
Moreover, we were unable to perform skin biopsy on HV with pruritus to compare the distribution of IENF in their skin with that of IBD patients.

Conclusion
The results of the present study suggest that IBD patients have dry skin features and frequently develop pruritus. These symptoms need to be recognized as clinical complications in IBD patients.