Journal Logo

Original Article

Significance of Erythema Nodosum and Pyoderma Gangrenosum in Inflammatory Bowel Diseases

A Cohort Study of 2402 Patients

Farhi, David; Cosnes, Jacques; Zizi, Nada; Chosidow, Olivier; Seksik, Philippe; Beaugerie, Laurent; Aractingi, Selim*; Khosrotehrani, Kiarash*

Author Information
doi: 10.1097/MD.0b013e318187cc9c
  • Free



Inflammatory bowel diseases (IBD) are due to an inappropriate and chronic activation of the intestinal immune system28. Major forms of IBD are Crohn disease and ulcerative colitis. Crohn disease is characterized by a chronic relapsing segmental transmural granulomatous inflammation that may involve any part of the digestive tract from mouth to anus, although mostly found in the ileum, the cecum, and the colon31. In contrast, ulcerative colitis is characterized by a nongranulomatous inflammation involving the intestinal mucosa from rectum to right colon in a continuous pattern, without participation of the rest of the bowel. Both diseases are multifactorial, and it has been suggested that the marked interpatient phenotypic variance may result from genetic and environmental heterogeneity.

IBD are associated with a broad spectrum of extraintestinal manifestations with wide ranges of reported incidence: 5%-20% for articular (peripheral or axial arthritis)4,26, 5%-15% for skin9,15, and 3%-10% for eye involvement4,5 (mainly episcleritis and uveitis). The pathogenesis of these associated manifestations has not been elucidated, although similar immune mechanisms have been suggested.

IBD-associated dermatologic manifestations are polymorphous, sometimes painful or disabling, and may herald the disease7,11. Therefore, they have important implications in the management of IBD. The most frequent skin manifestations include erythema nodosum (EN), aphthae, and neutrophilic dermatoses-chiefly pyoderma gangrenosum (PG)4,17. Perianal lesions have been reported in 14%-54% of patients with Crohn disease, and may present as large ulcerations, skin tags, fissures, fistulas, and abscesses29. However, in most cases they originate from the anus or the lower part of the rectum, and thus may not be considered genuine dermatologic manifestations.

Large cohort studies focusing on dermatologic involvement of IBD are scarce. Thus, it is not known whether the occurrence of EN and PG is associated with the clinical features and the prognosis of the intestinal manifestations. We conducted a descriptive study in a cohort of 2402 consecutive patients with IBD. Our primary objective was to assess the rate of EN and PG in patients with IBD. Second, we asked whether EN or PG was associated with any demographic, clinical, or prognostic factor in IBD.


General Design and Setting

Between 1975 and 2006, 5563 patients were referred for IBD to our academic tertiary referral gastroenterology department. We conducted a descriptive study of the skin manifestations among a cohort of 2402 consecutive patients between January 2000 and December 2005. This short time frame of 6 years was chosen to preserve homogeneity of clinical assessment and therapeutic management within the cohort.


All patients with IBD were included. They were classified, according to clinical, endoscopic, radiologic, surgical, and histologic cumulated data, into 3 groups: 1) Crohn disease; 2) ulcerative colitis; and 3) unclassified IBD. The term "unclassified IBD" has been recently proposed for patients with evidence of IBD, but without definitive histologic or other evidence to favor either Crohn disease or ulcerative colitis30.

Data Management

Demographic, clinical, and severity-related data of all patients with IBD were prospectively and systematically collected in a computerized database by 1 of the authors (JC) since 1994. Demographic data included age (at onset, at diagnosis, and at first referral to our center), sex, ethnic origin, professional status, and place of residency. Clinical data included family history, IBD phenotype (initial and cumulated location, surgical complications), classification of Crohn disease according to Montreal criteria33, and extraintestinal manifestations. Other factors such as body mass index, tobacco use (active or past smoking, cumulated pack-years, past maximum cigarettes/day), and hormonal factors (use of oral contraceptive, menopausal status) were also prospectively collected.

Assessment of IBD Severity

The clinical severity and aggressiveness of IBD are difficult to appraise reliably with a unique variable, as a single consensual severity score has not yet been proposed33. Therefore, the severity of IBD was approached as a composite concept. We used the following surrogate criteria: occurrence of surgical complications (perforation, stenosis); need for surgical intestinal resection33; cumulated value of the Post-Surgical Handicap Index (PSHI)10; requirement of a permanent stoma; occurrence of an intestinal cancer; IBD-related death; and need for systemic steroids, immunosuppressive agents, or artificial feeding33.

Dermatologic Manifestations

From 2000 to 2005, the presence of at least 1 cutaneous manifestation was prospectively reported in the database, and complete dermatologic data were recorded in the medical file. In December 2006, 3 dermatologists (NZ, KK, and DF) retrospectively collected this information from medical files for all patients prospectively coded as having at least 1 cutaneous manifestation. Hence, the precise dermatologic diagnosis was then added to the database. For the aim of the current study, only definite dermatologic diagnoses, based on typical clinical and/or histologic features, were retained17,34. We elected to focus on EN and PG since 1) these have been the most frequently reported skin diseases in IBD and 2) their clinical and histologic features are familiar to the caring physicians (gastroenterologists and dermatologists following patients of the cohort), and therefore classification bias was minimized. We also report data on aphthae, even though it is not clear whether these manifestations can be considered skin disorders or simply related to the digestive disease.

EN is defined by raised, erythematous or bluish-purple, subcutaneous, tender, multiple, and bilateral nodules measuring 1-5 cm in diameter, and usually located on the extensor surfaces of lower limbs, particularly on the pretibial area. Lesions undergo typical color changes, ending in temporary bruise-like areas, and then subsiding without scarring. PG is a painful neutrophilic dermatosis, featuring a rapidly extending ulceration, with a pustular and undermined border that is highly evocative: the edge is usually well limited and has a geometric-rather oval-shape; the internal slope that overhangs the ulcer is sharp, erythematous, often painful, and declines gradually over 1 to several cm12. PG occurs mostly on limbs, particularly on the pretibial area8.

For both EN and PG, the histopathologic examination of a lesional skin biopsy is not specific and is mainly useful in ruling out differential diagnoses in clinically atypical cases. In EN, it shows a neutrophilic perivascular reaction with septal panniculitis in the deep dermis and subcutaneous tissue, with a normal overlying epidermis19. In PG, histopathology reveals a dense neutrophilic dermal infiltrate, sometimes forming abscesses, and often associated with a perivascular lymphocytic infiltrate and a fibrinoid necrosis of the dermal vessel wall6. Histopathology was inconstantly performed in our patients to confirm the diagnoses of EN and PG. In addition, PG lesions in most cases were not biopsied since there might be a risk for aggravation. However, microbiologic analyses were routinely obtained from skin swabbing to rule out soft tissues infections.

When reviewing the patients' files, we considered the diagnoses of EN or PG as definitive when either the recorded clinical features were typical, or the clinical features were compatible and the histopathology was confirmatory.

Statistical Analysis

Statistical analysis was conducted in 3 stages. First, we described the demographic and clinical features of the cohort, overall and by major diagnostic groups (Crohn disease vs. ulcerative colitis). Second, we evaluated the frequency of EN and PG in our cohort, overall and by major diagnostic groups (Crohn disease vs. ulcerative colitis). Finally, we evaluated demographic, clinical, and severity-related factors associated with the most frequent dermatologic manifestations. Statistical analysis was performed with SAS 9.1 software (SAS Institute Inc., Cary, NC).

In univariate analyses, quantitative variables were compared with the Student test or the Mann-Whitney test, according to statistical conditions; qualitative variables were analyzed with the chi-square test or the Fisher exact test, according to statistical conditions. To take into account all potential confounding factors, all variables associated with a significance threshold ≤0.10 in the univariate analysis were included in the multivariate regression models.

In multivariate analyses, logistic regression models were implemented for the most frequent dermatologic manifestations.

Prespecified subgroup analysis was implemented for 3 of the studied factors: 1) hormonal factors (oral contraceptives, menopause) were assessed only in the subgroup of women; 2) Montreal classification criteria were assessed only in the subgroup of patients with Crohn disease; 3) PSHI score was assessed only in the subgroup of patients who underwent surgery.

All tests were 2-tailed. A p value ≤0.05 was defined as the significance threshold. Odds ratios (OR) were estimated with their 95% confidence intervals (95% CI).


Patient Characteristics

Of the 2402 included patients, 140 (5.8%) had at least 1 skin manifestation according to the computerized database, and their medical records were thoroughly checked for dermatologic data. Among the 2402 analyzed patients, 1521 had Crohn disease (63.3%), 744 had ulcerative colitis (31.0%), and 111 had unclassified IBD (4.6%).

Patients with Crohn disease were significantly younger than patients with ulcerative colitis, both at the first sign of disease (26.8 vs. 32.5 yr, respectively) and at the time of diagnosis (28.2 vs. 33.1 yr, respectively) (p < 0.0001 for all). Mean follow-up was 17.4 months. Median interval between first sign of IBD and diagnosis was 14.2 months. Female:male sex ratio was 1.55 (1.59 for Crohn disease vs. 1.18 for ulcerative colitis, p < 0.001). Most patients were white European individuals (71.6%) living in Paris and its suburbs (71.3%). There was no significant association between the type of IBD (Crohn disease vs. ulcerative colitis) and ethnic origin, profession, or current place of residence.

Clinical features are summarized in Table 1. A family history of IBD was present in 13.4% of patients, and there was a significant association between the familial type of IBD and the index case type of IBD (p < 0.0001). Most of the patients were between 17 and 40 years old (A2: 72.0%). The most frequent location of Crohn disease was terminal ileum (L1: 33.8%). The most frequent clinical type was inflammatory (B1: 48.6%).

Demographic and Clinical Data of 2402 Consecutive Patients With IBD Referred Between 2000 and 2005

Intestinal surgery for complications was required in 37% of the patients overall. The most common surgical complication was perforation, which mainly involved the perianal area. The mean cumulated PSHI score for the cohort was 26.1 ± 11.3 (range, 2-65). The need for systemic steroids (30.7% of the cohort), immunosuppressive agents (48.2%), and artificial feeding (6.5%) was significantly more frequent in patients with Crohn disease than in those with ulcerative colitis (p < 0.02).

The most common extraintestinal manifestations were articular involvement (14.6%), followed by dermatologic (5.8%), oral (5.8%), and ocular (3.1%) involvement. These manifestations were significantly more frequent in patients with Crohn disease than in patients with ulcerative colitis (p < 0.02 for all). The most frequent dermatologic manifestations were EN (4.0%) and PG (0.75%).

Time Between First Sign of IBD and a Dermatologic Manifestation

Among patients with skin manifestations, the median delay between the first digestive sign of IBD and the occurrence of the first dermatologic manifestation was 3.9 years, with a trend for an earlier occurrence of dermatologic manifestations in patients with Crohn disease compared with those with ulcerative colitis (3.5 vs. 5.9 yr, p = 0.13) for EN, but not for PG.

Relationship Between EN or PG and Clinical Features of IBD

Erythema Nodosum

EN occurred in 97 of 2402 patients (4.0%), and was significantly associated with Crohn disease (5.6% vs. 1.2% in ulcerative colitis, p < 0.0001) (Figure 1). The mean delay between the first digestive sign and the occurrence of EN was 3.9 years, without significant differences between Crohn disease and ulcerative colitis. EN preceded the digestive disease in 15 patients (15.5%), occurred 3 months to 21 years before in 6 patients (6.2%), and occurred simultaneously in 9 patients (9.3%).

Cumulative prevalence of erythema nodosum among 2402 patients with IBD, overall (A) and according to the disease group (B). In Figure 1A, the continuous line represents the cumulative prevalence of erythema nodosum and the dotted lines represent the 95% confidence interval.

Factors associated with EN in univariate analysis are shown in Table 2. EN did not show any significant association with ethnic origin, profession, place of residence, or familial history of IBD.

Univariate Analysis of the Association Between EN and Demographic, Clinical, and Severity-Related Factors in 2402 Consecutive Patients With IBD

Multivariate analyses of the factors associated with EN were conducted in 3 stages. First, factors associated with EN were assessed in the entire cohort (n = 2402). In this model (Table 3), EN was significantly and independently associated with a diagnosis of Crohn disease (p < 0.001), female sex (p < 0.001), eye (p < 0.001) and joint (p < 0.0001) involvement, and PG (p < 0.0001).

Multivariate Analysis of Factors Associated With EN

Second, among the factors of the Montreal classification, those associated with p ≤ 0.10 in univariate analysis (namely disease location and behavior criteria) were assessed for their association with EN in the subgroup of patients with Crohn disease (n = 1521) (Table 3). In this second model, EN was significantly and independently associated with isolated colonic involvement (L2 vs. L1: OR = 4.7 [2.0-10.8]; p = 0.0001), but not with the behavior criterion.

Third, hormonal factors associated with EN were assessed in the subgroup of women with a diagnosis of IBD (n = 1411). In this third multivariate model, none of the hormonal factors (oral contraception, menopause) was significantly associated with EN.

Pyoderma Gangrenosum

PG occurred in 18 of 2402 patients (0.75%), with no significant difference between the incidence in patients with Crohn disease and those with ulcerative colitis (Figure 2). A biopsy was performed in 4 cases that confirmed the clinical diagnosis. The mean delay between the first digestive sign and the occurrence of PG was 5.8 years, without significant differences between Crohn disease and ulcerative colitis. PG preceded the digestive diagnosis in 4 cases (22.2%), occurred 9.5 months before the first digestive sign of IBD in 1 patient (5.6%), and occurred simultaneously in 3 patients (16.6%). Overall, EN occurred significantly earlier than PG (3.9 yr after diagnosis vs. 5.8 yr; p < 0.05).

Cumulative prevalence of pyoderma gangrenosum among 2402 patients with IBD, overall (A) and according to the disease group (B). In Figure 2A, the continuous line represents the cumulative prevalence of pyoderma gangrenosum and the dotted lines represent the 95% confidence interval.

Factors associated with PG in univariate analysis are shown in Table 4. These were not related to diagnosis type and other personal factors, such as smoking or hormone (estroprogestative) treatment. Patients with PG more often needed a surgical treatment or immunosuppressive therapy for their IBD compared to patients without PG. Similar to EN, PG did not seem to make the diagnosis of the IBD easier or quicker: the delay to diagnosis was similar in groups of patients with and without PG.

Univariate Analysis of the Association Between PG and Demographic, Clinical, and Severity-Related Factors in 2402 Consecutive Patients With IBD

Multivariate analyses of factors associated with PG were conducted in 3 stages. First, factors associated with PG were assessed among all patients with IBD (n = 2402). In this model (Table 5), PG was significantly and independently associated with black African origin (p = 0.003), familial history of ulcerative colitis (p = 0.0005), uninterrupted pancolitis as the initial location of IBD (p = 0.03), permanent stoma (p = 0.002), eye involvement (p = 0.001), and EN (p < 0.0001). Among the 5 patients who had PG and permanent stoma, 2 had peristomal PG. Interestingly, even after exclusion of these 2 patients with peristomal PG, definitive stoma was more frequent in patients with PG than in patients without PG (18.8% vs. 2.5%, p = 0.07).

Multivariate Analysis of Factors Associated With PG

Second, the association between PG and the Montreal classification criteria was assessed in the subgroup of patients with Crohn disease (n = 1511) (Table 5). In this model, none of the Montreal criteria was significantly associated with PG.

Third, the association between PG and the PSHI score was assessed among the subgroup of patients with Crohn disease that underwent surgical intestinal resection (n = 536). In this subgroup, higher PSHI scores were not associated with PG.


We performed an analysis for aphthae, another common mucosal condition associated with IBD. Aphthae are defined by painful ulcerations most often occurring in the oral mucosa but that may also affect genital and anal mucosa as well. Aphthae were not considered a cutaneous manifestation, since they may be directly related to the digestive disease. Aphthae occurred in 31 of 2402 patients (1.3%), more frequently in patients with Crohn disease than in patients with ulcerative colitis (1.7% vs. 0.54%, respectively; p = 0.02). Multivariate analysis of factors associated with aphthae was conducted in 2 stages. Among all patients with IBD (n = 2402), aphthae were significantly and independently associated with joint involvement (p = 0.01), EN (p < 0.0001), perineal Crohn disease (p = 0.03), and PG (p = 0.0003). In the subgroup of patients with Crohn disease (n = 1521), the initial location of Crohn disease was not significantly associated with aphthae.


IBD are associated with a broad spectrum of manifestations with a variable degree of severity. Cutaneous signs are among the most frequent extraintestinal manifestations. We assessed the association between skin manifestations and digestive phenotype and severity in patients with IBD. We found that 1) EN and other extraintestinal manifestations-but not PG-are more frequent in Crohn disease than in ulcerative colitis; 2) EN and PG manifestations were associated with colonic IBD, as reported previously27; 3) most importantly, none of these manifestations was associated with any of the severity-related factors, apart from an association between permanent stoma and PG. The large size, systematic prospective data collection, and transversal cross-sectional design of the current study allow for minimizing both selection and classification bias. The adjustment on a large panel of demographic, clinical, and severity-related variables also reduces residual confounding bias. Moreover, our study population was comparable with previous large cohort studies for sex, age at onset, and criteria of the Montreal classification4,20,25,33. However, in most cases our patients did not undergo skin biopsy to confirm the proposed diagnosis of EN or PG, since in our routine practice this is not required when clinical features are typical.

In a literature review of skin manifestations in IBD, we found few good quality large-scale studies evaluating the frequency of dermatologic manifestations and the associated factors in IBD. We note that there is significant variation in reported rates of dermatologic manifestations and associated factors, reflecting the heterogeneity of the source-populations, recruitment flows, data collection processes, and diagnostic criteria. For obvious sensitivity reasons, reported rates are usually higher in prospective than in retrospective studies.

Previous studies4,23 analyzed the relationship between clinical features of IBD and extraintestinal manifestations altogether, rather than individually. When planning the current study, we identified no clinically or physiopathologically relevant grounds to support this design. Thus, we separately analyzed the relationship between features of IBD and each of the 2 major dermatologic manifestations. In addition, because we focused on the most frequent skin manifestations, which are easily recognizable by dermatologists as well as by gastroenterologists, we believe that our study does not underestimate the frequency of dermatologic signs. Therefore, our findings bear clearer practical information for the clinician managing IBD.

For most extraintestinal manifestations, the prevalence observed in the current cohort was similar to that reported previously. EN occurred in 4.0% of the patients with IBD, with higher rates in patients with Crohn disease. The preferential occurrence of EN in Crohn disease (rather than ulcerative colitis) has been previously reported, but is controversial5,9,15,18,21,25,26,34-36. This entity is the most commonly reported and the best studied dermatologic manifestation in IBD27, with rates ranging from 1% to 15%4,5,9,13,15,18,21-23,25-27,36. In the current series, PG occurred in 0.75% of 2402 consecutive patients with IBD. Reported incidence of PG varies between 0.70% and 1.3%5,13,22,23. These comparable figures between our cohort and previous reports indicate the validity of our data collection. EN was not associated with any ethnic group, and PG was independently associated with African American origin in contrast with previous reports by Nguyen et al25. They found a higher prevalence of EN among Hispanics compared to non-Hispanic whites and African Americans (10.7%, 3.4%, and 1.7%, respectively; p < 0.01). The prevalence of PG was comparable across these ethnic groups (2.1% overall)25. In our study population Hispanic groups were a minority, potentially explaining the differences with previous reports.

Smoking has been demonstrated to be 1 of the major risk factors for IBD. In the current study, even when we compared groups of long-term heavy smokers with nonsmokers, we did not find the previously described association between tobacco and dermatologic signs4,20. In a series of 173 consecutive Crohn disease patients, Barreiro-de Acosta et al4 found a positive association between smoking habits and EN (multivariate OR, 8.3; 95% CI, 1.1-66.1; p < 0.05). However, this was not reported for other extraintestinal manifestations. Manguso et al20 examined the relationship between tobacco use and extraintestinal manifestations among 535 consecutive patients with ulcerative colitis. They found increased dermatologic manifestations in smokers in univariate analysis (p = 0.001). These inconsistencies potentially result from differences in cohort populations where smoking may not have the same importance. In our population, more than 50% of patients were current or past smokers. This high prevalence of tobacco use may have shadowed its influence between groups of patients with or without extraintestinal manifestations. In a prospective study of 566 patients with Crohn disease and 271 patients with ulcerative colitis, Mendoza et al21 found a nonsignificant trend for a higher prevalence of PG in ulcerative colitis patients (1.5% vs. 1.0%, respectively, p > 0.05). In the current study, we did not find any significant relation between PG and ulcerative colitis.

Several authors5,13,16,22,25 have reported higher rates of extraintestinal-particularly dermatologic-manifestations among women with IBD, raising the hypothesis of an estrogen-mediated inflammatory process in the pathogenesis of these manifestations. In the current study, EN was significantly more frequent among women than it is in the general population14,19. Although in univariate analyses, estrogen treatment tended to be associated with EN and menopause was significantly associated with the absence of EN, none of these associations remained significant in multivariate analysis. Given the power of our statistical comparisons, we believe our study clearly shows the lack of association between estrogens and EN in IBD.

The occurrence of dermatologic signs in IBD patients always raises the question of a more severe course of the intestinal disease. It has been suggested that severe skin lesions might be the major indication for intestinal surgery since Monsen et al23 reported that PG healed in 5 patients after panproctocolectomy. In the current study, we assessed the relationship between dermatologic signs and digestive manifestations reflecting severe IBD. None of the dermatologic signs was independently and significantly associated with the severity of the intestinal involvement of IBD. In particular, dermatologic signs were not associated with surgical treatment or intense immunosuppressive regimens. In a study22 of 1132 patients with ulcerative colitis, the disease was active in 81% of patients with PG. In another cohort of 1274 patients with ulcerative colitis, PG was related with the activity of ulcerative colitis in 67% of cases23. In the current series, there was no significant association between presence of PG and aggressiveness of the IBD.

The occurrence of peristomal PG is usually considered a manifestation of pathergy-a peculiar trend to develop PG lesions at the site of trauma, either accidental or iatrogenic6. It is noteworthy that in the current study, patients with stoma tended to have more PG, even after the exclusion of patients with peristomal PG. However, because of the small number of patients with stoma who had PG, we must consider this result with caution.

We believe our results, in addition to describing potential associations within the cohort, provide important information on the pathophysiology of the extraintestinal manifestations in IBD. It is striking that very different symptoms, such as articular, eye, or skin involvement, occur within the same group of patients affected by IBD. This is witnessed by the strong associations observed in multivariate analyses. We believe this may be the translation of 2 phenomenons. First, the immune alterations leading to the digestive disease could affect other epithelia if immune cells such as lymphocytes or neutrophils have the capacity to home to these tissues. Accordingly, aberrant homing of neutrophils has been reported in PG1,32. One could hypothesize that beside the skin, other tissues such as cartilage or eye mucosa can also be affected by similar integrin alterations, explaining the strong associations between the various extracutaneous manifestations. The second phenomenon evoked by this strong association is genetic susceptibility. It is tempting to speculate that some patients who have IBD may have a genetic susceptibility to develop extraintestinal manifestations if they bear allelic variants favoring, for example, altered homing of lymphocytes or neutrophils. Of note, although CARD15 mutations do not account for the occurrence of PG in Crohn disease24, familial forms of PG have been reported2,3. Accordingly, EN has been associated with CARD15 polymorphisms in Crohn disease patients4.

In conclusion, in the current study we demonstrate that dermatologic manifestations are not associated with a more aggressive intestinal disease. The association between PG and stoma should be more carefully investigated in a larger prospective cohort. Overall, clinicians should be encouraged not to modify the IBD treatment regimen based on the skin manifestations. In addition, the presence of any of these signs should lead to a search for perianal disease or other extraintestinal involvement, including joint, mouth, and eye manifestations. In the current study we provide estimates of the prevalence of the most frequent dermatologic manifestations in IBD, and appraise their relationship with demographic data and disease phenotype. Better knowledge of the associations between dermatologic manifestations and the phenotype of IBD may provide clinicians with additional accuracy in the diagnosis and management of IBD.


We thank Florentia Kaguelidou, MD, for assistance in manuscript presentation and Marie-France Guillaume for assistance in data collection.


1. Adachi Y, Kindzelskii AL, Cookingham G, Shaya S, Moore EC, Todd RF III, Petty HR. Aberrant neutrophil trafficking and metabolic oscillations in severe pyoderma gangrenosum. J Invest Dermatol. 1998;111:259-268.
2. al-Rimawi HS, Abuekteish FM, Daoud AS, Oboosi MM. Familial pyoderma gangrenosum presenting in infancy. Eur J Pediatr. 1996;155:759-762.
3. Alberts JH, Sams HH, Miller JL, King Lee Jr. Familial ulcerative pyoderma gangrenosum: a report of 2 kindred. Cutis. 2002;69:427-430.
4. Barreiro-de Acosta M, Dominguez-Munoz JE, Nunez-Pardo de Vera MC, Lozano-Leon A, Lorenzo A, Pena S. Relationship between clinical features of Crohn's disease and the risk of developing extraintestinal manifestations. Eur J Gastroenterol Hepatol. 2007;19:73-78.
5. Bernstein CN, Blanchard JF, Rawsthorne P, Yu N. The prevalence of extraintestinal diseases in inflammatory bowel disease: a population-based study. Am J Gastroenterol. 2001;96:1116-1122.
6. Brooklyn T, Dunnill G, Probert C. Diagnosis and treatment of pyoderma gangrenosum. BMJ. 2006;333:181-184.
7. Burgdorf W. Cutaneous manifestations of Crohn's disease. J Am Acad Dermatol. 1981;5:689-695.
8. Callen JP. Pyoderma gangrenosum. Lancet. 1998;351:581-585.
9. Christodoulou DK, Katsanos KH, Kitsanou M, Stergiopoulou C, Hatzis J, Tsianos EV. Frequency of extraintestinal manifestations in patients with inflammatory bowel disease in Northwest Greece and review of the literature. Dig Liver Dis. 2002;34:781-786.
10. Cosnes J, de Parades V, Carbonnel F, Beaugerie L, Ngo Y, Gendre JP, Sezeur A, Gallot D, Malafosse M, le Quintrec Y. Classification of the sequelae of bowel resection for Crohn's disease. Br J Surg. 1994;81:1627-1631.
11. Crowson AN, Nuovo GJ, Mihm MC Jr, Magro C. Cutaneous manifestations of Crohn's disease, its spectrum, and its pathogenesis: intracellular consensus bacterial 16S rRNA is associated with the gastrointestinal but not the cutaneous manifestations of Crohn's disease. Hum Pathol. 2003;34:1185-1192.
12. Farhi D. The clinical and histopathological description of geometric phagedenism (pyoderma gangrenosum) by Louis Brocq one century ago. Arch Derm. 2008;144:75.
13. Freeman HJ. Erythema nodosum and pyoderma gangrenosum in 50 patients with Crohn's disease. Can J Gastroenterol. 2005;19:603-606.
14. Garcia-Porrua C, Gonzalez-Gay MA, Vazquez-Caruncho M, Lopez-Lazaro L, Lueiro M, Fernandez ML, Alvarez-Ferreira J, Pujol RM. Erythema nodosum: etiologic and predictive factors in a defined population. Arthritis Rheum. 2000;43:584-592.
15. Greenstein AJ, Janowitz HD, Sachar DB. The extra-intestinal complications of Crohn's disease and ulcerative colitis: a study of 700 patients. Medicine (Baltimore). 1976;55:401-412.
16. Lakatos PL, Szalay F, Tulassay Z, Molnar T, Kovacs A, Gasztonyi B, Papp J, Lakatos L. Clinical presentation of Crohn's disease. Association between familial disease, smoking, disease phenotype, extraintestinal manifestations and need for surgery. Hepatogastroenterology. 2005;52:817-822.
17. Lebwohl M, Lebwohl O. Cutaneous manifestations of inflammatory bowel disease. Inflamm Bowel Dis. 1998;4:142-148.
18. Lindgren A, Wallerstedt S, Olsson R. Prevalence of Crohn's disease andsimultaneous occurrence of extraintestinal complications and cancer. An epidemiologic study in adults. Scand J Gastroenterol. 1996;31:74-78.
19. Mana J, Marcoval J. Erythema nodosum. Clin Dermatol. 2007;25:288-294.
20. Manguso F, Sanges M, Staiano T, Gargiulo S, Nastro P, Gargano D, Somma P, Mansueto G, Peluso R, Scarpa R, D'Armiento FP, Astarita C, Ayala F, Renda A, Mazzacca G, D'Arienzo A. Cigarette smoking and appendectomy are risk factors for extraintestinal manifestations in ulcerative colitis. Am J Gastroenterol. 2004;99:327-334.
21. Mendoza JL, Lana R, Taxonera C, Alba C, Izquierdo S, Diaz-Rubio M. [Extraintestinal manifestations in inflammatory bowel disease: differences between Crohn's disease and ulcerative colitis.] Med Clin (Barc). 2005;125:297-300.
22. Mir-Madjlessi SH, Taylor JS, Farmer RG. Clinical course and evolution of erythema nodosum and pyoderma gangrenosum in chronic ulcerative colitis: a study of 42 patients. Am J Gastroenterol. 1985;80:615-620.
23. Monsen U, Sorstad J, Hellers G, Johansson C. Extracolonic diagnoses in ulcerative colitis: an epidemiological study. Am J Gastroenterol. 1990;85:711-716.
24. Newman B, Cescon D, Domenchini A, Siminovitch KA. CD2BP1 and CARD15 mutations are not associated with pyoderma gangrenosum in patients with inflammatory bowel disease. J Invest Dermatol. 2004;122:1054-1056.
25. Nguyen GC, Torres EA, Regueiro M, Bromfield G, Bitton A, Stempak J, Dassopoulos T, Schumm P, Gregory FJ, Griffiths AM, Hanauer SB, Hanson J, Harris ML, Kane SV, Orkwis HK, Lahaie R, Oliva-Hemper M, Pare P, Wild GE, Rioux JD, Yang H, Duerr RH, Cho JH, Steinhart AH, Brant SR, Silverberg MS. Inflammatory bowel disease characteristics among African Americans, Hispanics, and non-Hispanic Whites: characterization of a large North American cohort. Am J Gastroenterol. 2006;101:1012-1023.
26. Orchard TR, Wordsworth BP, Jewell DP. Peripheral arthropathies in inflammatory bowel disease: their articular distribution and natural history. Gut. 1998;42:387-391.
27. Orchard TR, Chua CN, Ahmad T, Cheng H, Welsh KI, Jewell DP. Uveitis and erythema nodosum in inflammatory bowel disease: clinical features and the role of HLA genes. Gastroenterology. 2002;123:714-718.
28. Podolsky DK. Inflammatory bowel disease. N Engl J Med. 2002;347:417-429.
29. Sandborn WJ, Fazio VW, Feagan BG, Hanauer SB. AGA technical review on perianal Crohn's disease. Gastroenterology. 2003;125:1508-1530.
30. Satsangi J, Silverberg MS, Vermeire S, Colombel JF. The Montreal classification of inflammatory bowel disease: controversies, consensus, and implications. Gut. 2006;55:749-753.
31. Shanahan F. Crohn's disease. Lancet. 2002;359:62-69.
32. Shaya S, Kindzelskii AL, Minor J, Moore EC, Todd RF III, Petty HR. Aberrant integrin (CR4; alpha(x)beta2; CD11c/CD18) oscillations on neutrophils in a mild form of pyoderma gangrenosum. J Invest Dermatol. 1998;111:154-158.
33. Silverberg MS, Satsangi J, Ahmad T, Arnott ID, Bernstein CN, Brant SR, Caprili R, Colombel JF, Gasche C, Geboes K, Jewell DP, Karban A, Loftus EV Jr, Pene AS, Riddel RH, Schar DB, Schreiber S, Steinhart AH, Targan SR, Vermeire S, Warren BF. Toward an integrated clinical, molecular and serological classification of inflammatory bowel disease: report of a Working Party of the 2005 Montreal World Congress of Gastroenterology. Can J Gastroenterol. 2005;19(Suppl A):5-36.
34. Trost LB, McDonnell JK. Important cutaneous manifestations of inflammatory bowel disease. Postgrad Med J. 2005;81:580-585.
35. Turkcapar N, Toruner M, Soykan I, Aydintug OT, Cetinkaya H, Duzgan N, Ozden A, Duman M. The prevalence of extraintestinal manifestations and HLA association in patients with inflammatory bowel disease. Rheumatol Int. 2006;26:663-668.
36. Veloso FT, Carvalho J, Magro F. Immune-related systemic manifestations of inflammatory bowel disease. A prospective study of 792 patients. J Clin Gastroenterol. 1996;23:29-34.
© 2008 Lippincott Williams & Wilkins, Inc.