Oral Diseases (2015) 21, e98–e104 doi:10.1111/odi.12236 © 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd All rights reserved www.wiley.com

ORIGINAL ARTICLE

Characterisation of a Swedish cohort with orofacial granulomatosis with or without Crohn’s disease 2 € G Gale1, S Ostman , E Rekabdar3,  A Torinsson Naluai3, K H€ogkil4, B Hasseus1, R Saalman5*, 1 M Jontell * 1

Department of Oral Medicine and Oral Pathology, Institute of Odontology, The Sahlgrenska Academy, University of Gothenburg, G€ oteborg; 2Department of Infectious Diseases, The Sahlgrenska Academy, University of Gothenburg, G€ oteborg; 3Genomics Core Facility, The Sahlgrenska Academy, University of Gothenburg, G€ oteborg; 4Eastman Institute, Public Dental Health, Stockholm; 5 Department of Paediatrics, Institute of Clinical Sciences, The Sahlgrenska Academy, University of Gothenburg, G€ oteborg, Sweden

OBJECTIVE: To compare oral manifestations in a Swedish cohort of patients with orofacial granulomatosis with or without Crohn0 s disease and to assess NOD2 polymorphisms in the two groups. METHODS: Twenty-nine patients with orofacial granulomatosis were included. Demographics, disease history, clinical features and concurrent Crohn0 s disease were recorded. DNA was extracted from buccal swabs and examined for NOD2 variants Arg702Trp, Gly908Arg and Leu1007fsinsC, all previously linked to gastrointestinal Crohn0 s disease. RESULTS: Twelve of 29 patients were diagnosed with coexisting gastrointestinal Crohn0 s disease, and of whom 21 were males. Symptom duration was significantly longer for the orofacial granulomatosis group compared to the group with coexisting Crohn0 s disease (P < 0.0001). The orofacial granulomatosis patients also perceived their overall discomfort, aesthetic problems and social discomfort as more severe. No significant differences in the clinical presentation of oral lesions between the two groups were found. None of the patients with orofacial granulomatosis carried any of the NOD2 variations, whereas four of the 12 patients with coexisting Crohn0 s disease had a NOD2 variant (Arg702Trp). CONCLUSION: The two patient groups had similar phenotypic characteristics but seemed to have genotypic differences regarding NOD2. The Swedish cohort differed in their clinical characteristics from patients reported in other geographical regions. Oral Diseases (2015) 21, e98–e104

Correspondence: Mats Jontell, Department of Oral Medicine and Pathology, Institute of Odontology, The Sahlgrenska Academy, University of Gothenburg, PO Box 450, 405 30 G€oteborg, Sweden. Tel: +46 31 786 31 33, Fax: +46 31 824733, E-mail: [email protected] *Saalman and Jontell have contributed equally as senior investigators. Received 20 January 2014; revised 3 March 2014; accepted 5 March 2014

Keywords: mucosal diseases; inflammatory bowel disease; NOD2; clinical characterisation; oral mucosa

Introduction Orofacial granulomatosis (OFG) is an orofacial disorder with clinical features characterised by lip swelling, mucosal tag formation, cobblestone phenomenon and angular cheilitis. Non-caseating granulomas with multinucleate giant cells, lymphoedema and perivascular lymphocytic infiltration dominate the histopathology (Wiesenfeld et al, 1985; Freysdottir et al, 2007; Patel et al, 2010). However, granulomas are not always present in the biopsies, and a general infiltration of inflammatory cells may be the only histopathological finding (Hegarty et al, 2003; Tilakaratne et al, 2008; McCartan et al, 2011). OFG may appear alone or in combination with systemic diseases such as Crohn’s disease (Wiesenfeld et al, 1985; Saalman et al, 2009) and sarcoidosis (Takada et al, 1993). Available scientific data do not point to a single factor that can explain the development of OFG in all patients who suffer from this disease (Giovannetti et al, 2012). Thus, the development of OFG most likely represents the outcome of interacting factors including genetic susceptibility (Stosiek et al, 1987; Gibson and Wray, 2000), environmental factors such as the microflora (Ivanyi et al, 1993; Gibson et al, 2000; Apaydin et al, 2004; Savage et al, 2004) and host immune responses (Lim et al, 1997; Freysdottir et al, 2007; Patel et al, 2010; Giovannetti et al, 2012). Genetic susceptibility has previously been studied in patients with CD, and nucleotide oligomerisation domain 2 (NOD2) has been found to be strongly associated with this disease (Hugot et al, 2001; Ogura et al, 2001). No similar studies have been published focusing on patients with OFG, despite its obvious association with CD. Studies of demographic and clinical characteristics in cohorts of patients with OFG stem mainly from the UK

OFG in a Swedish cohort

G Gale et al

and Ireland (Wiesenfeld et al, 1985; Armstrong et al, 1997; Al Johani et al, 2010; Campbell et al, 2011; McCartan et al, 2011). Within the UK, the majority of cases appear to be concentrated in Scotland. Whether this may be a consequence of various environmental factors and/or a genetic susceptibility has yet to be elucidated. Thus, it is important to characterise the phenotype and genotype of OFG cohorts from different geographical regions to investigate how various factors may contribute to the aetiology of this condition. The primary objective of this study was to evaluate whether any demographic and/or clinical characteristics, with focus on the specific oral features, may distinguish patients with OFG solely from those with coexisting CD. In addition, the presence of NOD2 variations was determined in this Swedish OFG cohort.

Patients and methods Patients At Sahlgrenska Academy, University of Gothenburg, the Department of Oral Medicine and Pathology has collaborated with the Department of Pediatrics for the last decade on oral manifestations of inflammatory bowel diseases in both paediatric and adult patients. This collaboration has resulted in the establishment of a national resource for patients with OFG and a multidisciplinary network involving microbiologists, allergy specialists, immunologists and geneticists. In this study, a total of 29 patients were included from 2003 to 2012. The cohort was divided into two categories, one presenting with OFG solely (OFG-S; n = 17) and the other with concurrent CD (OFG + CD; n = 12). The diagnosis of OFG was established by an experienced specialist in oral medicine (MJ) and was based primarily on clinical characteristics (Wiesenfeld et al, 1985). All except one patient displayed a swollen lip or facial swelling at the primary visit. The patient without lip or facial swelling had CD and presented with focal nodular and erythematous gingiva containing granulomas in the oral biopsy. In all other patients, at least one of the following characteristics was observed in addition to swollen lip or facial swelling: mucosal tags, cobblestone phenomenon (‘basket of eggs’ appearance) or gingival enlargement with an erythematous nodular appearance. Representative incisional biopsies of intraoral lesions were performed on the majority of the patients. Furthermore, all patients were carefully examined for possible systemic diseases, particularly inflammatory disorders in the gastrointestinal tract. A diagnosis of CD was confirmed on the basis of the Porto criteria (Escher et al, 2005) using endoscopic, radiological and histological investigations. An endoscopic investigation was carried out in all patients, who had a clinical history of gastrointestinal symptoms and/or had increased inflammatory parameters in blood and faeces (sedimentation rate, C-reactive protein, faecal calprotectin) at the time of inclusion or developed such parameters during the follow-up period. None of the patients in this cohort were diagnosed with Melkersson–Rosenthal syndrome, sarcoidosis or confirmed food-induced OFG.

All data, including clinical photographs, were recorded in a structured electronic form and then prepared for the registration (Jontell et al, 2005). Sites and types of oral manifestations as well as perceived complaints of orofacial lesions were all registered. The reported complaints were recorded as the following: overall discomfort, aesthetic problems, social constraints and oral pain, for example related to food intake. The patients were asked to express their perceived complaints using a visual analogue scale (VAS with score markers at each increment of 1).

e99

Analyses of NOD2 variants Analyses of available allele and/or genotype frequencies for NOD2 variants Arg702Trp, Gly908Arg and Leu1007fsinsC, all linked to CD, were performed. Buccal epithelial cells were collected using Isohelix SK1 buccal swabs. The buccal swabs were sent to Eppendorf AG, Hamburg, Germany, for DNA extraction according to standard techniques. PCR and genetic analyses were carried out at the Genomic Core Facilities, Sahlgrenska Academy, University of Gothenburg, Gothenburg. Amplification performed by touch-down PCR (TD-PCR) was performed in GeneAmpâ PCR System 9700 (Life technologies, www.lifetechnologies. com) in 5 ll reaction mixture containing 2.5 ll AmpliTaq Goldâ 360 Master Mix (Life technologies), 0.4 ll of each primer (forward and reverse) (10 lM) and 20 ng genomic DNA and 0.7 ll H2O. The primers used for the PCR are shown in Table 1, and the ExonPrimer (http://ihg.gsf.de/ ihg/ExonPrimer.html) was used to design primers to cover the CDS region of the exons. The PCR products were purified using magnetic beads (Ampure, Agencourt, Biomekâ FX, Beckman Coulter, www.beckmancoulter.com) in the automated workstation Biomekâ FX (Beckman Coulter) and eluted in 15 ll H2O. Sequence-PCR, using BigDyeâ Terminator v 3.1 Cycle Sequence Kit (Life Technologies), was carried out in a GeneAmpâ PCR System 9700 (Life Technologies) according to standard procedures in 10 ll reaction mixture containing 6 ll template (TD-PCR product), 0.25 ll BDT v.3.1, 1.875 ll BDT buffer (5X) and 1 ll primer (1.6 lM) (forward or reverse) for each region or exon. The sequence-PCR products were purified using magnetic beads (CleanSeq, Agencourtâ, Biomekâ FX, Beckman Coulter) in the automated workstation Biomekâ NX Table 1 Primers used for PCR amplification

Primer

Primer length (bp)

Sequence 50 -30

EX4-1F EX4-1R EX4-2F EX4-2R EX4-3F EX4-3R EX4-4F EX4-4R EX8-F EX8-R EX11-F EX11-R

20 21 18 20 17 18 18 20 20 20 23 18

CAGCCCATTGTCTGGTTAGG GAACTTGAACTCGTCAAAGCC GCTGTTGGCCTGATGTTG ACCTTGGGAAGCTGAGTCTG GGGTCCCCAAAGACCAC GTCTGGCACTCAGCCAGC GAAAGGACAGCAGCGTGG CTCGGTGCTCCCACACTTAG CTGCAGAGGGAGGAGGACTG AAGACCTTCAGAACTGGCCC AACTCCTGCAGTCTCTTTAACTG CTGCCATTCCTCTCTCCC

Oral Diseases

OFG in a Swedish cohort

G Gale et al

e100

(Beckman Coulter) and eluted in high-dye formamide. The products were loaded on a 3730 DNA Analyzer (Life Technologies), and the results were analysed using the software programs Sequencing Analysis (v. 5.2; Applied Biosystems, Life technologies) and SeqScape (v.2.5; Applied Biosystems, Life technologies). Studies with available allele and/or genotype frequencies for NOD2 variants Arg702Trp, Gly908Arg and Leu1007fsinsC, all linked to CD, were included. Statistical analyses Statistical analysis was conducted using the nonparametric Mann–Whitney U-test or Fisher’s exact test (presence of NOD2 mutation) (GraphPad Prism; GraphPad Software, La Jolla, CA, USA). P-values 48 3 25–48 >48 25–48 25–48 25–48 >48 25–48 4 25–48 4 25–48 1 2 12 1 12 2 1 2

Lip swelling

Facial swelling

Tag formation

Cobblestone phenomenon

X

X X

X X

X X

X

X X X X

X

X X X X X X X X X X X X X X X X X X X X X X X X X

X X X X X X X X X X X X X X X X X X X X X

X X X

X X X X X X X X X X

Angular cheilitis

Oral Diseases

Ulcer

Gingival lesion

NOD2 variants

X X

X X

X X X

X

X

X X X X X X X X X X

X X X X X

X X X

X

X

X

X

X

X

X

X

X

X

X X

X X

X X

X

X X

X

X X

X X X

The duration of symptoms was based on retrospective data reported by the patients at the primary examination. OFG-S: orofacial granulomatosis solely. c OFG + CD; orofacial granulomatosis with coexisting Crohn’s disease. b

Perioral erythema

X

X X

X

OFG in a Swedish cohort

G Gale et al

1:2.6. The median age at inclusion was 14 years with no significant differences between the two groups. At the primary oral examination, the patients were asked to specify the time point of onset of the orofacial symptoms. Fifteen of the 17 patients with OFG-S were diagnosed more than a year after the first appearance of symptoms, and three patients had lived with their symptoms for more than 5 years prior to diagnosis (Table 2). In contrast, the time period from onset of symptoms to established OFG diagnosis was significantly shorter in the OFG + CD group (P < 0.001). Six OFG patients had a first-degree relative with autoimmune disease. In the OFG-S group, one relative had diabetes, one had CD and one had ankylosing spondylitis, whereas in the OFG + CD group, two relatives had diabetes and one had psoriatic arthritis. None of the OFG patients themselves presented with such autoimmune diseases. Three patients in the OFG-S group, but none in the OFG + CD group, reported that they had first-degree relatives with OFG symptoms. Twelve of 29 OFG patients reported that they had a first-degree relative who suffered from allergy, seven patients in the OFG-S group and five patients in the OFG + CD group. Regarding allergy of all patients in the study, nine patients had a clinical history of food allergy (fish, shellfish, cow’s milk, cacao, nuts and/or peanuts) with symptoms such as anaphylaxis, acute angioedema, vomiting, diarrhoea and/or nausea (five in the OFG-S and four in the OFG + CD group). Seven patients reported that they suffered from allergy induced by airborne allergens, predominantly rhinoconjunctivitis (four in the OFGS and three in the OFG + CD group). Further, eight of 29 patients reported that they suffered from eczema, six in the OFG-S group and two in the OFG + CD group. Additionally, three patients in the OFG-S group suffered from asthma, but in these cases, there was no clear relation to a specific allergen confirmed. No asthma was reported from the group of patients with OFG + CD. Self-reported complaints All patients were asked to report various aspects of their perceived orofacial complaints using a structured form including a VAS. The OFG-S patients reported significantly higher levels of overall discomfort, at the peak of their disease than did the OFG + CD patients (OFG-S vs OFG + CD; P < 0.05). The reported aesthetic problems (OFG-S vs OFG + CD; P < 0.001) and social discomfort (OFG-S vs OFG + CD; P < 0.05) were also more pronounced in the OFG-S group compared with the OFG + CD group. These problems resulted in an unwillingness to participate in leisure activities and a strong desire to avoid taunts at school. Both groups reported equal levels of oral pain induced by food intake. For example, oral pain was reported when the swollen lips became fissured with open ulceration. Sites of orofacial components at the primary examination The majority of OFG-S patients had lip involvement (94%), followed by involvement of the buccal mucosa (77%), sulcus (71%) and facial region (76%) (Figure 1).

Similar distribution was seen among the OFG + CD patients, but here, the most frequently involved sites were the sulcus (83%) and lips (83%) followed by the buccal mucosa (67%). The gingivae were affected in one third of the patients. Only two patients showed manifestations in the floor of the mouth (staghorn appearance), and none of the patients had any lesions in the palatal mucosa. Thirteen of the 29 patients (45%) presented with angular cheilitis. There were no significant differences between the two OFG groups for any of the variables regarding specific distribution or the number of sites involved.

e101

The spectrum of orofacial features at the primary examination The main oral clinical features in the patient cohort described in Table 2 and Figure 2 were swelling affecting the lips (OFG-S/OFG + CD; 83% vs 83%), mucosal tag formation (OFG-S/OFG + CD; 71% vs 75%; P = ns), cobblestone phenomenon (OFG-S/OFG + CD; 71% vs 42%; P = ns) and facial swelling (OFG-S/OFG + CD; 76% vs 92%). Tag formation was registered when the buccal mucosa showed multiple tags and when the sulcus displayed linear mucosal ridges. Cobblestone phenomenon was recorded when the buccal mucosa presented with a ‘basket of eggs’ appearance. Angular cheilitis was observed in almost half of the patients (OFG-S/ OFG + CD; 47% vs 42%; P = ns) and a minority presented with perioral erythema (OFG-S/OFG + CD; 35% vs 25%; P = ns). Ulceration was the most rare of all clinical features observed in both groups (OFG-S/OFG + CD; 24% vs 8%; P = ns). The presence of NOD2 variations in the OFG cohort Variation in NOD2 has been shown to predispose for CD, and this was the reason we decided to analyse NOD2 polymorphisms in our cohort. None of the 17 patients with OFG-S had any of the NOD2 variants previously shown to be associated with CD. In contrast, four of 12 of our OFG + CD patients carried a single copy of the NOD2 variation (Arg702Trp variant; Table 2). The four patients with NOD2 variations did not differ in disease severity compared with the rest of the OFG-CD patients. Although the patient number was limited, univariate data analysis revealed a significant difference in NOD2 variation between OFG-S and OFG + CD patients (P < 0.05). Neither of the NOD2 variants Gly908Arg nor Leu1007fsinsC was found in any of the patient populations. Principal component analysis and discriminant analysis Principal component analysis and discriminant analysis (PCA-DA) were used to assess how multivariate data contribute to a possible discrimination between the groups of OFG-S and OFG + CD. We compiled a number of variables including demographic data, clinical characteristics, NOD mutations, heredity for immunoregulatory diseases, clinical history of allergy, autoimmunity and self-reported complaints of OFG (X-variables in PCA-DA). The diagnoses OGF-S and OFG + CD served as Y-variables. As shown in the score scatter plot in Figure 3a, both the

Oral Diseases

OFG in a Swedish cohort

G Gale et al

e102

(a)

Figure 1 Lesion sites in patients with orofacial granulomatosis solely (OFG-S) and orofacial granulomatosis with coexisting Crohn’s disease (OFG + CD).

(b)

Figure 2 Mucosal textures in patients with orofacial granulomatosis solely (OFG-S) and orofacial granulomatosis with coexisting Crohn’s disease (OFG + CD).

patients with OFG-S and patients with OFG + CD tend to form clusters, and consequently, the two groups became separated from each other. The column-loading plot, shown in Figure 3b, depicts the X-variables that contributes most to this separation. The closer the X-variable is to Y and the higher the bar and the lower the spread, the stronger the positive correlation. Thus, the following X-parameters were more associated with OFG-S patients than with OFG + CD patients: perceived overall discomfort, aesthetic problems and social complaints. The dominant X-parameter linked to OFG + CD was the presence of NOD2 variants.

Discussion In this study, we have investigated a Swedish cohort of patients with OFG for oral phenotype and genotype. All patients were diagnosed with OFG during the course of the study. We found no significant differences in the clinical presentation of oral lesions between those who had only OFG and those who also suffered from CD. However, our data indicate that the groups experienced their symptoms differently and that there were differences in genotype, that is, in NOD2 variants. Oral Diseases

Figure 3 Principal component analysis and discriminant analysis (PCADA). (a) Score scatter plot and (b) column loading plot depicting the association between clinical diagnosis, that is, orofacial granulomatosis solely (OFG-S) or orofacial granulomatosis with coexisting Crohn’s disease (OFG + CD) and X-variables, including demographic and oral clinical features. In (a), each dot represents one individual and individuals with similar characteristics (X-variables) cluster, whereas individuals that differ separate. In (b), X-variables with bars projected in the same direction as OFG-S or OFG + CD are positively associated, whereas bars in the opposite direction are inversely related to the diagnosis. The larger the bar and the smaller the error bar, the stronger and more certain is the contribution to the model. R2Y indicates how well the variation of Y is explained, while Q2 indicates how well Y can be predicted. The closer these values are to 1, the better the model. In this model, R2Y = 0.78, Q2 = 0.34.

OFG in a Swedish cohort

G Gale et al

The present study was designed in part to enable a detailed comparison with previously published results from other geographical areas such as the UK and Ireland (Campbell et al, 2011; McCartan et al, 2011). In this Swedish cohort of patients with OFG with or without Crohn0 s disease, we found no significant difference in the oral phenotype. This was true both for the site of the oral lesions and for the type of the oral manifestations. In contrast, a recent study from the UK reported distinct oral clinical features in each of the two groups (Campbell et al, 2011). Involvement of the buccal sulcus, ulcers and mucosal scarring was all found to be more prominent in the OFG + CD group, while the OFG-S group was reported to have more lip involvement. Furthermore, in our study, no significant differences in median age or gender distribution were found between the OFG-S patients and the OFG + CD patients. However, in comparison with reported cohorts from the UK and Ireland, the patients were younger with a median age of 14 years (range 7–32 years) than those reported from the UK (23 years, 7–73 years) (Campbell et al, 2011) and Ireland (28 years, 5–84 years) (McCartan et al, 2011). The gender distribution also differed markedly, with a clear predominance of males (female/male ratio 1:2.6) in our study, whereas the patients from the UK and Ireland showed a more equal distribution. These differences between the cohorts in terms of age and gender are unlikely to be due to biased recruitment of patients. The patients included in our study are representative of those we have seen in recent decades. Thus, our results suggest that Swedish OFG patients debut at a much younger age and showed a different gender distribution compared to OFG patients from the UK and Ireland. There was also a clear difference between the UK cohort and our Swedish cohort in terms of the oral sites involved, in both the OFG-S and OFG + CD groups. The lips and buccal mucosae were the most affected sites in both studies, but there were obvious differences between the occurrence of changes in the sulcus region and the gingiva. Campbell and co-workers (Campbell et al, 2011) found far more gingival changes than we have observed (63.5% vs 35%) while we found noticeably more changes in the sulcus region in our population (15% vs 76%). The Irish study (McCartan et al, 2011), which included patients with OFG solely, did not give any detailed information regarding changes in the sulcus region, but the gingival involvement was calculated as 35%, which was comparable with our observations. The frequency of tag formation was similar in both of our OFG-S and OFG + CD patients (72% vs 75%), but this was much higher than reports from the paediatric population from the UK (35%) (Campbell et al, 2011) and the Irish OFG patients (30%) (McCartan et al, 2011). The high prevalence of tags in our cohort may be partly explained by the young age of our patients as a correlation between age and tag formation has been reported from the UK, that is, paediatric patients in the UK have been found to show tag formations more commonly than adult patients (Campbell et al, 2011). When our two patient groups were compared, the OFGS patients reported their complaints significantly higher on

the VAS than the patients in the OFG + CD group did. The main complaint was related to the disfigurement caused by the swollen lips, which is in accordance with a recent study by McCartan and co-workers (McCartan et al, 2011). One explanation to the reported differences regarding complaints may be that the OFG-S group had a more extended duration from the onset to the time of diagnosis than the OFG + CD patients. Another possible explanation for the discrepancy in the VAS score may be that the symptoms from the gastrointestinal tract in the OFG + CD group may have overshadowed the oral symptoms in some of the cases. A high proportion of our patients with OFG, whether they had Crohn0 s disease or not, reported that they had allergies. This is in line with a recent report from the UK where a positive history of allergies was observed in about 80% of the OFG patients with or without Crohn0 s disease (Patel et al, 2013). However, in this Swedish cohort, we did not find a positive history of allergies to as high an extent as the UK report. There has been debate over whether OFG-S and OFG + CD represent different disease entities or whether they signify the same disease with different involvement of the gastrointestinal tract (Savage et al, 2004; Sanderson et al, 2005; Zbar et al, 2012). NOD2 was the first gene associated with CD (Hugot et al, 2001; Ogura et al, 2001), and since then, over 70 genes or loci have been associated with CD (Cleynen et al, 2013), but the NOD2 gene still has the strongest linkage. This strong linkage prompted us to specifically investigate NOD2 variants in our cohort of OFG patients. Interestingly, our data support the view that the NOD2 gene variation pattern may differ between those with only OFG and those with OFG + CD. None of the 17 patients in our cohort with OFG-S carried any CD-linked NOD2 variation. In contrast, four of 12 patients with OFG + CD displayed a NOD2 variation. Thus, although our study comprises a limited number of patients, it is the first to support the notion that OFG-S and OFG + CD may be genetically different. This finding needs to be confirmed in a larger cohort of patients with OFG. A limitation of our present study is that the number of patients was not as large as those in the studies from the UK and Ireland. However, our group represents a wellknown healthcare facility for OFG patients that have been nationally recognised by physicians and dentists. This has most likely ensured that the majority of patients in Sweden diagnosed during the course of the study have been included. Although the patient number is limited, we feel that it is important to present data from our geographical region. Another limitation might be that the patients with OFG-S may represent a heterogeneous group as one could not exclude that they may later go on to develop CD. This is an uncertainty, which is difficult to account for, but the median time of our follow-up time was 3 years and never

Characterisation of a Swedish cohort with orofacial granulomatosis with or without Crohn's disease.

To compare oral manifestations in a Swedish cohort of patients with orofacial granulomatosis with or without Crohn's disease and to assess NOD2 polymo...
389KB Sizes 0 Downloads 2 Views