Secondary Logo

Journal Logo

Systemic Vasculitis: A Cause of Indeterminate Intestinal Inflammation

Brogan, P. A.*; Malik, M.; Shah, N.*; Kilday, J. P.*; Ramsay, A.*; Shah, V.*; Murch, S. H.; Thomson, M. A.; Walker-Smith, J. A.; Lindley, K. J.*; Milla, P J.*; Dillon, M. J.*

Journal of Pediatric Gastroenterology and Nutrition: April 2006 - Volume 42 - Issue 4 - p 405-415
doi: 10.1097/01.mpg.0000215305.63417.26
Original Articles: Gastroenterology
Free

Objectives: Indeterminate intestinal inflammation may result from a variety of inflammatory conditions in addition to ulcerative colitis and Crohn disease. The primary systemic vasculitides may present with intestinal inflammation and an indeterminate colitis. We set out to describe a series of children with primary systemic vasculitis who initially presented with clinical features suggestive of inflammatory bowel disease (IBD) to establish criteria that might help discriminate between IBD and primary systemic vasculitis.

Methods: Ten children (6 boys, median age at presentation 8.9 years, range 0.9-14.5 years) satisfied inclusion criteria.

Results: All had abdominal pain, weight loss, diarrhea (6 of 10 bloody) and laboratory evidence of a severe acute phase response. Extraintestinal clinical features included vasculitic rash, renal impairment, myalgia, testicular pain and polyarthritis. Endoscopy showed vascular changes or other macroscopic findings suggestive of vasculitis in 5 of 10 patients. Gut histology revealed indeterminate chronic inflammatory mucosal changes and one patient with small artery fibrinoid necrosis in the submucosal vessels. Extraintestinal biopsy was performed in 6 patients and had a higher yield for the demonstration of vasculitis than intestinal biopsy. The results of selective visceral angiography was suggestive of vasculitis in all patients, but was normal in 7 cases of treatment-unresponsive classic IBD. Treatment comprised corticosteroid and azathioprine in all patients. Cyclophosphamide was given to 7 of 10 patients.

Conclusions: Extraintestinal manifestations and inflammatory responses that may be disproportionate to the degree of intestinal inflammation provide clues to the presence of an underlying primary systemic vasculitis, and these data suggest that selective visceral angiography plays a key role in the diagnosis of vasculitis in this context. It is important to identify and treat any vasculitic component because failure to do so may result in consequential morbidity or mortality.

Abbreviations: IBD, inflammatory bowel disease-UC, ulcerative colitis-CD, Crohn disease-ANCA, antineutrophil cytoplasmic antibodies-PAN, polyarteritis nodosa-ACR, American College of Rheumatology-ENA, extractable nuclear antigens.

*Institute of Child Health and Great Ormond Street Hospital and †Royal Free Hospital, London, UK

Received July 29, 2005; accepted February 8, 2006.

Address for correspondence and reprint request to Dr. P. A. Brogan, Clinical Lecturer in Vasculitis, BSc (Hon), MBChB (Hon), MRCPCH, MSc, PhD, Department of Rheumatology, Institute of Child Health and Great Ormond St Hospital for Children, 30 Guilford St, London WC1N 1EH, UK. (e-mail: P.Brogan@ich.ucl.ac.uk).

Supported in part by the Charlotte Parkinson Research Fund.

Back to Top | Article Outline

INTRODUCTION

In young people, a variety of inflammatory disorders may present with inflammatory bowel disease (IBD) with a histologically indeterminate colitis. Some will ultimately declare themselves as either ulcerative colitis (UC) or Crohn disease (CD). Others such as the primary systemic vasculitides occasionally mimic IBD (1-18), and extraintestinal vasculitis can complicate IBD (19-45). Thus, the differentiation between primary systemic vasculitis and IBD can be clinically challenging, but it is important to distinguish these disorders because their treatment and outcomes are different. More importantly, an IBD-like presentation of primary vasculitis in adult life may be extremely severe, with fatal outcome in many cases (46,47).

Primary systemic vasculitis often presents insidiously and sequentially, affecting different organ systems at different times. Thus, the presentation to different specialists with protean clinical manifestations virtually always creates uncertainty, and in our collective experience, the diagnosis is rarely, if ever, straightforward in vasculitis with predominantly gastrointestinal involvement as the initial presenting feature.

Previous studies have examined the hypothesis that the granulomata of CD in fact represent granulomatous vasculitis and that the disease is a mesenteric vasculitic process resulting in inflammatory microvascular occlusion and subsequent multifocal gastrointestinal infarction (48,49). It has also been suggested that such a pathogenetic mechanism would explain many of the intestinal and extraintestinal manifestations of CD (48,49).

Vasculitic extraintestinal complications of IBD have been reported in adults, but less so in children. These include vasculitis of the retina (26,28,36), brain (19,23,30,34), skin (20-22,24,33,39,40,45,50,51), muscle (25,29,44), joints (25,42,52) and lung (27,31,37,53). Moreover, the observation of antineutrophil cytoplasmic antibodies (ANCA) in some patients with UC and CD (usually perinuclear and directed against mixed epitopes) further lends credence to the hypothesis that a small vessel vasculitis could be a prominent pathogenetic mechanism in IBD (38,43,54,55).

The aims of this study were to describe a series of children who satisfied the classification criteria for primary systemic vasculitis but who initially presented with symptomatic intestinal inflammation mimicking IBD. Clinical, histological, radiological and immunologic features that might help discriminate between IBD and primary systemic vasculitis were sought.

Back to Top | Article Outline

PATIENTS AND METHODS

Retrospective review of case notes of children presenting to 2 paediatric tertiary referral centers (Great Ormond Street Hospital for Children and the Royal Free Hospital) in London, UK, between 1993 and 2002 was performed. Patients were identified from a computer database and clinicians' records of children with vasculitis. Inclusion criteria were children with a diagnosis of vasculitis (based on clinical, angiographic and biopsy features) subsequently classified as polyarteritis nodosa(PAN) with the use of the American College of Rheumatology criteria (56), in whom the major initial presentation was symptomatic intestinal inflammation, suggesting a differential diagnosis of IBD (UC or CD). Patient demography, clinical features, colitis scores (57), immunology, histology, radiology, treatment and outcomes were examined. Exclusion criteria were those with a proven primary or acquired immunodeficiency, or malignancy. Z-scores for height and weight were calculated and analysed against normative age-related data with the use of a one-sample t test. The Mann-Whitney U test was used to compare acute phase responses between the vasculitis patients (n = 10) and an IBD control group (n = 7).

Back to Top | Article Outline

RESULTS

Demography, Clinical Features and Colitis Scores

The initial presenting features and investigative results of all the patients are summarized in Table 1. Ten children (6 boys, mean age at presentation 8.9 years, range 0.9-14.5 years) satisfied the inclusion criteria. The median time to diagnosis of vasculitis was 24 months (range 2-120 months), with most patients having a chronic and sequential presentation of their illness and with intestinal disease as the predominant symptom at initial presentation. All had abdominal pain, failure to thrive (median height Z-score −1.2, P = 0.04; median weight Z-score −1.1, P = 0.02), diarrhoea (6 of 10 bloody) and laboratory evidence of a severe acute phase response with elevation of erythrocyte sedimentation rate (ESR) (median 72 mm/h, range 56-190), C reactive protein (CRP) (median 72 mg/L, range 17-100), thrombocytosis (median platelet count 487 × 109/L, range 342-1013) and hypoalbuminaemia (mean 33 g/L, range 8-42). The mean colitis score (as defined by Beattie et al. (57)) was 4.5/10 (range 3-7/10). For comparison, the acute phase markers of 7 children with IBD but with normal visceral arteriography results (see succeeding sections and Table 2) are listed in Table 3. ESR and CRP were higher in the vasculitis patients, although it should be noted that there were some missing data points in the IBD group, thus limiting statistical power (particularly for comparison of CRP and platelet counts).

TABLE 1

TABLE 1

TABLE 2

TABLE 2

TABLE 3

TABLE 3

Back to Top | Article Outline

Extraintestinal Manifestations

Two of 10 patients had moderate to severe renal impairment (glomerular filtration rate less than 60 mL/min/1.73m2), with glomerular filtration rates of 50 mL/min/1.73m2 (case 1) and less than 10 mL/min/1.73m2 (case 10), the latter requiring renal replacement therapy. The remaining patients had a calculated glomerular filtration rate greater than 100 mL/min/1.73m2. The median plasma creatinine level for the overall group was 62 μmol/L (range 21-1460 μmol/L). Three of 10 patients had microscopic haematuria, and 2 of 10 had both haematuria and proteinuria (urine albumin:creatinine ratio, 0.55 and 2.9 mg/mg; reference range <0.1 mg/mg). These latter 2 patients also had both haematuria and proteinuria on dipstick and abnormal renal biopsy results (see succeeding sections).

Rash was present in 7 of 10 patients. Three had a petechio-purpuric rash (Fig. 1A); one had an eczematoid rash with vasculitic nail changes (Fig. 1B); one had a vesicular rash in addition to purpura; one had subcutaneous nodules and one had a diffuse ecchymotic swelling of the hand (Fig. 1C), in addition to pretibial erythema nodosum. Nonerosive polyarthritis was present in 6 of 10 patients and predominantly affected the knees, ankles and hips (Fig. 1D). Symmetrical and asymmetrical joint involvement was observed. Other extraintestinal vasculitic features included myalgia (6 of 10), and testicular pain and tenderness (a sign of testicular vasculitis, 1 of 6 boys).

FIG. 1

FIG. 1

In 3 cases (cases 3, 8 and 9; Table 1), there was no evidence of a normal physiological sinus arrhythmia on examination, with the electrocardiogram showing a fixed R-R interval. In 2 of these cases, the patients experienced life-threatening massive gastrointestinal haemorrhage, the severity of the blood loss masked by a failed tachycardic response, with no increase in heart rate before collapsing with hypovolaemic shock.

Extraintestinal features in the 7 children with IBD but with normal arteriography results were less evident (Table 2). One child had a rash of uncertain vasculitic nature (did not undergo biopsy). One child had an episode of noninfectious pericarditis before the onset of CD.

Back to Top | Article Outline

Immunologic Features

The presence of perinuclear ANCA (pANCA) was detected using indirect immunofluorescence (IIF) in 2 of 10 patients. At that time, it was not routine to perform enzyme-linked immunoabsorbent assay (ELISA) for ANCA type, although this was performed in patient 10 and showed that the ANCA was against myeloperoxidase. Antienterocyte antibodies detected by IIF were weakly positive in 2 of 10 patients. Three of 10 patients had anticardiolipin (ACL) antibody positivity (IgG) [patient 2, ACL IgG 15.6 (0-12 arbitrary ELISA units); patient 8, ACL IgG 44.5; and patient 10, ACL IgG 23.4]. Other autoantibodies including antinuclear antibody, antibodies against extractable nuclear antigens (ENAs) and antiglomerular basement membrane antibodies were negative in all patients. Serum complement (C3 and C4) was normal in all. Serum immunoglobins were normal in 8 of 10 patients; one (case 1) had elevated IgA (3.49 g/L, reference range 0.8-2.8 g/L); one had borderline low IgM (0.4 g/L, reference range 0.5-2 g/L) of dubious significance. T-cell subsets were normal in all 10 patients. Results of hepatitis B and C serology were negative in all patients. The test for anti-Saccharomyces cerevisiae mannan antibodies was not available when most of the patients presented, so this was not examined in this series.

Back to Top | Article Outline

Radiological Findings

Results of selective visceral angiography were suggestive of vasculitis in all 10 patients, with involvement of the renal arteries in 8 of 10 patients and with mesenteric or hepatic arterial bed involvement in all 10 patients who had this investigation. We have previously documented a full description of the aneurysmal and nonaneurysmal arteriographic changes observed in childhood PAN (58), but important signs specifically sought included large and small aneurysms, perfusion defects, the presence of collateral arteries, lack of crossing of peripheral renal arteries and delayed emptying of small renal arteries (58). Angiographic changes suggestive of vasculitis affected exclusively the small- to medium-sized arteries (20%-50% of the caliber of the main renal artery (58)) in both renal and mesenteric arterial beds. Both aneurysmal and nonaneurysmal angiographic changes (58) were observed (Table 1) and are demonstrated inFigure 2A-B. Intestinal uptake of technetium 99m-labelled white cells was increased in 5 of 6 studies performed.

FIG. 2

FIG. 2

Visceral angiograms had been performed in an additional 7 cases with severe active IBD, whose clinical details are presented in Table 2. The procedures were performed to exclude a diagnosis of covert vasculitis because of failure to respond to standard medical therapy, with features possibly suggestive of autoimmunity (presence of autoantibodies in the absence of other obvious causes including chronic infection or malignancy), and/or failure to respond to conventional therapy for IBD as documented in Table 2. In all cases, the angiograms were entirely normal, confirming that severe acute IBD is not in itself sufficient to cause multifocal arteriographic changes.

Back to Top | Article Outline

Endoscopic and Histological Findings

All patients underwent upper and lower gastrointestinal endoscopy. In 5 of 10 patients, colonoscopy revealed discrete areas of mucosal ulceration, with arterial narrowing and haemorrhage around vessels clearly evident. The mucosa between lesions was endoscopically normal (Fig. 3). These appearances were atypical of either UC or CD.

FIG. 3

FIG. 3

Histological examination of the gastrointestinal mucosal biopsies revealed a variable picture. Findings in individual cases are summarized in Table 1. Pathological changes were focal and frequently involved only oneof several biopsies taken from the same patient. Only one patient showed an identifiable vasculitis in the submucosal bowel wall (case 10). In one patient (case 3), even when the full thickness of the resected bowel wall was examined, no overt vasculitis (ie, fibrinoid necrosis of vessels) was found. In another patient (case 6), no abnormalities were seen in endoscopic biopsies from both upper and lower gastrointestinal tract. In the remaining cases, the most common finding was a nonspecific increase in chronic inflammatory cells in the lamina propria (Fig. 4A and B), with an accompanying patchy acute inflammation. The acute component consisted of scattered neutrophils in the surface epithelium and glandular crypts leading to a focal, rather subtle, cryptitis. The latter was seen in stomach (1 patient), colon (3 patients) and rectum (2 patients). Submucosal chronic inflammation was identified in 1 endoscopic biopsy series (case 5) and 2 resection specimens (cases 3 and 10). Distortion of glandular architecture was seen in 2 patients. One patient showed only a mild increase in chronic inflammatory cells in the lamina propria without an acute component (case 1). One patient (case 3) showed a histological picture typical of CD in that both endoscopic biopsy and resection of sigmoid colon showed ulceration, glandular distortion, mucin depletion, crypt abscess formation, cryptitis and the presence of submucosal epithelioid granulomas.

FIG. 4

FIG. 4

Extraintestinal biopsy was performed in 6 patients. A total of 5 skin biopsies were performed; one showed a leukocytoclastic vasculitis, one showed granulomatous inflammation, two showed lymphocytic perivascular dermatitis (Fig. 4C) and one showed no pathology. A renal biopsy was carried out in case 1 and case 10. In both cases, there was a pauci-immune focal segmental glomerulonephritis with focal fibrin deposition, consistent with a vasculitis. A muscle biopsy in one case showed myositis but no overt vasculitis.

Back to Top | Article Outline

Treatment and Outcome

After the diagnosis of vasculitis, all patients received systemic corticosteroid therapy (starting dose of prednisolone 2 mg/kg orally for 4-8 weeks, subsequently weaning to a maintenance dose of 0.2-0.5 mg/kg on alternate days for a mean of 4.2 years, range 2-7.2 years), with at least one additional drug. Cyclophosphamide was given to 8 of 10 patients [orally at a dose of 2 mg/kg for 2-6 months in 4 patients, and in monthly intravenous pulses at 750 mg/m2 for 9 months in 2 patients; mean cumulative cyclophosphamide dose (oral and intravenous) 249 mg/kg, range 120-360 mg/kg]. In all 8 patients who received cyclophosphamide, previously intractable disease was brought into remission with this agent, although patient 10 went on to develop rapidly progressive renal failure 14 months after treatment and required dialysis and subsequent renal transplantation. Azathioprine was used in all 10 patients at an oral dose of 2 mg/kg once daily for a mean duration of 2.8 years (range 0.2-6 years). Other therapies included mycophenolate mofetil (2 of 10 patients), cyclosporin (1 patient), aminosalicylate derivatives (6 of 10 patients), colchicine (1 patient), plasma exchange (2 of 10 patients) and infliximab (2 of 10 patients).

All 10 patients have had a chronic relapsing course, with significant intestinal and extraintestinal morbidity. At mean follow-up of 4 years (range 2-7 years), 7 of 10 patients are currently in remission (Table 1; cases 1, 2, 4, 6, 8, 9 and 10), although case 10 required renal transplantation. Three of 10 patients have ongoing chronic symptomatic intestinal inflammation, with abdominal pain and intermittent bloody diarrhoea (cases 3, 5 and 7). Only 2 patients are off all therapy (cases 2 and 9).

Back to Top | Article Outline

DISCUSSION

A proportion of patients who present with IBD have on histology an indeterminate intestinal inflammation. In some, it will ultimately become clear that they have UC or CD. In others, the inflammatory changes may be due to another inflammatory disorder such as that associated with an immunodeficiency or a primary systemic vasculitis. As their clinical course, treatment and outcome may all be quite different, differentiation of the nature of the underlying inflammatory process is important. In particular, the course of the colitis of a primary vasculitis may be severe and often fatal (46,47). We describe a series of children who had chronic intestinal mucosal inflammation and systemic vasculitis in whom there was a degree of uncertainty as to whether the underlying diagnosis was that of a primary systemic vasculitis or a systemic vasculitis complicating a primary IBD such as UC or CD. In 8 cases, the disease was life threatening and did not respond to conventional immunosuppression for CD (Table 2), whereas cyclophosphamide induced long-term remission. The presence of vasculitis was defined in all patients on the basis of tissue biopsy and/or selective visceral angiography and in the presence of clinical features of vasculitis with no other cause. It is of interest that in 5 patients, macroscopic mucosal changes including small vessel abnormalities were seen at colonoscopy. Vasculitis was later confirmed at mesenteric angiography.

Although the patients satisfied the classification criteria for polyarteritis nodosa (56,59) (cases 1-9) or microscopic polyangiitis (case 10), it must be emphasized that such classification criteria are of limited specificity in adults (56,60) and have never been formally validated in children (61). The classification of the childhood vasculitides remains controversial, and no single system has yet proven to be entirely satisfactory (61). In fact, there has been a recent consensus meeting relating to the classification of paediatric vasculitis, which will be published in the near future (Dillon, personal communication, 2005). It is important to emphasize, however, that classification criteria are not the same as diagnostic criteria, although the former are often confused and misused as the latter (60). Classification criteria work best in the study of groups of patients and work less well in the diagnostic evaluation of individual patients. As an example, the American College of Rheumatology criteria for PAN were designed to differentiate PAN from other types of vasculitis but not to diagnose vasculitis in the first instance (60). In this article, we diagnosed systemic vasculitis on the basis of histology and/or arteriography and in the absence of other identifiable causes such as infection or malignancy. Following diagnosis, classification criteria as defined by the American College of Rheumatology (56) were then applied to define the vasculitis type that, on the whole, was polyarteritis nodosa in this series, although case 10 was best classified as microscopic polyangiitis as defined by the Chapel Hill Consensus criteria (59).

Gut histology was indeterminate for CD or UC and was of low sensitivity for the identification of vasculitis. Although pathological abnormalities were detected in all but one case, changes were often minor and were frequently restricted to a single region in one of several biopsies. The findings reflect the fact that tissue biopsy per se is of limited sensitivity for the diagnosis of vasculitis because the disease is patchy (62,63). Where a vasculitic process involves the bowel, the changes are predominantly submucosal (48). Routine endoscopic biopsies typically sample only mucosa and a very small portion of submucosa. Extraintestinal biopsy had a higher yield for the identification ofvasculitis in this series, but generally has limited sensitivity alone for the diagnosis of vasculitis (62,63). One limitation of these observations however is that, on the whole, tissue biopsy was performed after therapy with corticosteroids was initiated, potentially limiting the diagnostic sensitivity further.

Of prime importance is not whether the vasculitis is primary or secondary but the detection of its presence in the first instance because vasculitis (primary or reactive) is associated with significant morbidity and mortality if not treated in its own right (64). Indeed, therapy with cyclophosphamide, corticosteroid and plasma exchange can be life saving where there is fulminant vasculitic disease (cases 1, 9 and 10).

Antineutrophil cytoplasmic antibodies, although diagnostically helpful for the identification and subsequent classification of some small vessel vasculitides in adults, in isolation are poorly sensitive for the diagnosis of vasculitis in children (65). Moreover, they may not correlate with disease activity (66). Indeed, only 2 of the 10 patients in this series were found to be ANCA positive, although one limitation of this retrospective series was that ELISA was not routinely performed in all for the detection of ANCA.

Selective visceral angiography was suggestive of vasculitis in all 10, and we would suggest that this investigation plays a key role in the work-up of children with suspected vasculitis (58,62,63). It is important to appreciate, however, that both aneurysmal and nonaneurysmal angiographic changes are manifest in children with systemic vasculitis and that incorporation of the nonaneurysmal signs into definitions of angiogram positivity for vasculitis significantly increases the diagnostic sensitivity of the test (58,62,63), particularly in those children who may have received immunosuppression before having the test performed. Although labelled white cell scanning was useful for documenting the presence of intestinal inflammation, it did not discriminate between vasculitis and IBD.

We remain interested in the observation of a fixed R-R interval in 3 patients. Case 9 (Table 1) presented with recurrent massive gastrointestinal haemorrhage requiring cardiopulmonary resuscitation on one occasion, when the patient showed no increasing tachycardia or hypotension on quarter-hourly observations before collapsing with asystole. The electrocardiogram showed a fixed R-R interval, although echocardiography was unremarkable in these 3 cases. We speculate (but cannot confirm) that the compensatory tachycardia that should have accompanied such a degree of blood loss was blunted, perhaps as a result of small arterial vasculitis affecting the cardiac pacemaker.

A detailed account of the treatment of the childhood vasculitides is beyond the scope of this article, but induction therapy with oral corticosteroid, cyclophosphamide and antiplatelet therapy (aspirin or dipyridamole), followed by maintenance therapy with low-dose alternate-day prednisolone and a steroid sparing agent (usually azathioprine), is recommended (67). More invasive therapies such as high-dose pulsed intravenous methylprednisolone, plasma exchange or, more recently, biologic therapy are reserved for those who do not enter remission with standard therapy or who present with fulminant acutely life-threatening vasculitis (67) (as in cases 1 and 10 in this series). The potential benefit derived from the addition of biologic therapy such as infliximab as adjunctive treatment for those patients whose disease is recalcitrant despite cyclophosphamide (patient 10) or in whom there are concerns relating to cyclophosphamide toxicity from high cumulative doses remains unproven and anecdotal (68).

It is worthy of some comment that there was a considerable delay in the diagnosis of vasculitis in many of these patients, and hence, a delay in initiation of therapy with cyclophosphamide. This observation emphasizes the sequential way that many of the patients presented, with months of intestinal symptoms predating the onset of more florid and typical vasculitic features such as nephritis. We suggest that one important message from this article would be to promote awareness of vasculitis as a possibility in children with atypical intestinal inflammation and extraintestinal symptoms and that selective visceral angiography can be deployed earlier in cases where there is doubt regarding the diagnosis.

In conclusion, primary systemic vasculitis is a cause of indeterminate colitis and can mimic IBD in its clinical presentation. Extraintestinal manifestations and acute phase responses that may be disproportionate to the degree of intestinal mucosal inflammation, and endoscopic appearance provide clues to the presence of an underlying primary systemic vasculitis or a reactive vasculitis. Other clinical features such as active urinary sediment, vasculitic rash or presence of renal failure provide clues to the presence of an underlying vasculitis. Ultimately, however, selective visceral angiography incorporating the full spectrum of arteriographic changes observed in childhood PAN, combined with targeted tissue biopsy, plays a key role in the diagnosis of suspected vasculitis. It is important to identify and treat any vasculitic component (whether primary or reactive) in its own right because failure to do so may result in considerable morbidity or even mortality.

Back to Top | Article Outline

Acknowledgments:

This work was supported in part by the Charlotte Parkinson Research Fund. Conflict of interest: none.

Back to Top | Article Outline

REFERENCES

1. Aizawa Y, Komeji S, Irimajiri S, et al. Effective use of azathioprine in a case of polyarteritis nodosa with recurrent intestinal perforation: a distinction between polyarteritis nodosa and allergic granulomatous angiitis. Nippon Rinsho 1974;32:3634-9.
2. Akbarian M. Abdominal apoplexy in polyarteritis nodosa. Report of a case. Am J Dig Dis 1966;11:63-7.
3. Banke L. Polyarteritis nodosa in the alimentary canal. Ugeskr Laeger 1967;129:518-20.
4. Blau EB, Morris RF, Yunis EJ. Polyarteritis nodosa in older children. Pediatrics 1977;60:227-34.
5. Carron DB, Douglas AP. Steatorrhoea in vascular insufficiency of the small intestine. Five cases of polyarteritis nodosa and allied disorders. Q J Med 1965;34:331-40.
6. Castro-Salomo A, Pou E, Ordi J, et al. Gastro-intestinal bleeding due to granulomatous vasculitis. Clin Exp Rheumatol 1994;12:232.
7. Colton CL, Butler TJ. The surgical problem of polyarteritis nodosa. Br J Surg 1967;54:393-6.
8. Dahlander B, Odeberg H, Hansen BU. Polyarteritis nodosa: ischaemic intestinal pain successfully treated with nitroglycerin. J Intern Med 1996;240:31-2.
9. Elias N, Sabo E, Naschitz JE, et al. Colonic ulcers in a patient with hepatitis C virus-associated polyarteritis nodosa. J Clin Gastroenterol 1998;26:212-5.
10. Fontaney F, Dacosta A, Mosnier JF, et al. Inflammatory pseudotumor of the intestine caused by necrotizing vasculitis in the course of Horton disease. Gastroenterol Clin Biol 1995;19:940-3.
11. Freilich BL, Bernstein CN. Vasculitis possibly confined to the small and large intestine. West J Med 1995;162:63-5.
12. Gladman DD, Ross T, Richardson B, et al. Bowel involvement in systemic lupus erythematosus: Crohn's disease or lupus vasculitis? Arthritis Rheum 1985;28:466-70.
13. Gundogdu HZ, Kale G, Tanyel FC, et al. Intestinal perforation as an initial presentation of polyarteritis nodosa in an 8-year-old boy. J Pediatr Surg 1993;28:632-4.
14. Hunsicker RC, James PM Jr. Intestinal polyarteritis nodosa: a surgical impasse? South Med J 1972;65:337-40.
15. Karp DR, Kantor OS, Halverson JD, et al. Successful management of catastrophic gastrointestinal involvement in polyarteritis nodosa. Arthritis Rheum 1988;31:683-7.
16. Lee EL, Smith HJ, Miller GL, et al. Ischemic pseudomembranous colitis with perforation due to polyarteritis nodosa. Am J Gastroenterol 1984;79:35-8.
17. Mazy V, Delforge M, Demoulin JC, et al. Ischemic necrosis of the intestine: expression of a vasculitis. Rev Med Liege 1994;49: 412-21.
18. Nicks AJ, Hughes F. Polyarteritis nodosa "mimicking" eosinophilic gastroenteritis. Radiology 1975;116:53-4.
19. Adamek RJ, Wegener M, Wedmann B, et al. Cerebral vasculitis in Crohn disease. Leber Magen Darm 1993;23:91-3.
20. Allegue RF, Boixeda DM, Moreno IR, et al. Leukocytoclastic vasculitis and ulcerative colitis. Rev Esp Enferm Apar Dig 1988; 74:185-6.
21. Aram H, Rubinstein N, Granot E. Annular erythema with histologic features of leukocytoclastic vasculitis in ulcerative colitis. Cutis 1985;35:250-2.
22. Callen JP. Severe cutaneous vasculitis complicating ulcerative colitis. Arch Dermatol 1979;115:226-7.
23. Carmona MA, Jaume AF, Ramirez RJ. Cerebral thrombosis and vasculitis: an uncommon complication of ulcerative colitis. Bol Asoc Med P R 2000;92:9-11.
24. Chalvardjian A, Nethercott JR. Cutaneous granulomatous vasculitis associated with Crohn's disease. Cutis 1982;30:645-55.
25. Disdier P, Swiader L, Harle JR, et al. Crohn's disease and gastrocnemius vasculitis: two new cases. Am J Gastroenterol 1997;92:880-2.
26. Duker JS, Brown GC, Brooks L. Retinal vasculitis in Crohn's disease. Am J Ophthalmol 1987;103:664-8.
27. Forrest JA, Shearman DJ. Pulmonary vasculitis and ulcerative colitis. Am J Dig Dis 1975;20:482-6.
28. Garcia-Diaz M, Mira M, Nevado L, et al. Retinal vasculitis associated with Crohn's disease. Postgrad Med J 1995;71:170-2.
29. Gilliam JH III, Challa VR, Agudelo CA, et al. Vasculitis involving muscle associated with Crohn's colitis. Gastroenterology 1981;81:787-90.
30. Gobbele R, Reith W, Block F. Cerebral vasculitis as a concomitant neurological illness in Crohn's disease. Nervenarzt 2000;71:299-304.
31. Isenberg JI, Goldstein H, Korn AR, et al. Pulmonary vasculitis-an uncommon complication of ulcerative colitis. Report of a case. N Engl J Med 1968;279:1376-7.
32. Jenkins HR, Jewkes F, Vujanic GM. Systemic vasculitis complicating infantile autoimmune enteropathy. Arch Dis Child 1994;71:534-5.
33. Kay MH, Wyllie R. Cutaneous vasculitis as the initial manifestation of Crohn's disease in a pediatric patient. Am J Gastroenterol 1998;93:1014.
34. Nelson J, Barron MM, Riggs JE, et al. Cerebral vasculitis and ulcerative colitis. Neurology 1986;36:719-21.
35. Raby O, Salvanet-Bouccara A, Forestier F. Uveopapillitis and vasculitis in Crohn disease. Bull Soc Ophtalmol Fr 1987;87:1205-10.
36. Salmon JF, Ursell PG, Frith P. Neovascular glaucoma as a complication of retinal vasculitis in Crohn disease. Am J Ophthalmol 2000;130:528-30.
37. Sargent D, Sessions JT, Fairman RP. Pulmonary vasculitis complicating ulcerative colitis. South Med J 1985;78:624-5.
38. Satchell SC, Anthony P, Nicholls AJ. Systemic vasculitis and p-ANCA positivity developing in a patient with ulcerative colitis and the antiphospholipid syndrome. Nephron 1999;83:365-7.
39. Schlehaider UK, Suckow M, Rosenthal P, et al. Cutaneous reactions in Crohn disease. Vasculitis in various skin segments. Hautarzt 1997;48:328-31.
40. Slater DN, Waller PC, Reilly G. Cutaneous granulomatous vasculitis: presenting feature of Crohn's disease. J R Soc Med 1985;78:589-90.
41. Smith OP, Hann IM. Systemic vasculitis complicating infantile autoimmune enteropathy. Arch Dis Child 1995;72:467.
42. Speiser JC, Moore TL, Zuckner J. Ulcerative colitis with arthritis and vasculitis. Clin Rheumatol 1985;4:343-7.
43. Weir A, Taylor-Robinson SD, Poole S, et al. Cytoplasmic antineutrophil cytoplasmic antibody-positive vasculitis associated with ulcerative colitis. Am J Gastroenterol 1997;92:506-8.
44. Weizman Z. Vasculitis involving muscle associated with Crohn's colitis. Gastroenterology 1982;82:1483-4.
45. Zlatanic J, Fleisher M, Sasson M, et al. Crohn's disease and acute leukocytoclastic vasculitis of skin. Am J Gastroenterol 1996; 91:2410-3.
46. Levine SM, Hellman DB, Stone JH. Gastrointestinal involvement in polyarteritis nodosa (1986-2000): presentation and outcome in 24 patients. Am J Med 2002;112:386-91.
47. Case records of the Massachusetts General Hospital, Case 35082. New Engl J Med. 1949;240:308-10.
48. Wakefield AJ, Sawyerr AM, Dhillon AP, et al. Pathogenesis of Crohn's disease: multifocal gastrointestinal infarction. Lancet 1989;2:1057-62.
49. Wakefield AJ, Sankey EA, Dhillon AP, et al. Granulomatous vasculitis in Crohn's disease. Gastroenterology 1991;100:1279-87.
50. Castanet J, Lacour JP, Perrin C, et al. Cutaneous vasculitis with lesions mimicking Degos' disease and revealing Crohn's disease. Acta Derm Venereol 1995;75:408-9.
51. Verbov J, Stansfeld AG. Cutaneous polyarteritis nodosa and Crohn's disease. Trans St Johns Hosp Dermatol Soc 1972;58:261-8.
52. Wackers FJ, Tytgat GN, Vreeken J. Necrotizing vasculitis and ulcerative colitis. Br Med J 1974;4:83-4.
53. Collins WJ, Bendig DW, Taylor WF. Pulmonary vasculitis complicating childhood ulcerative colitis. Gastroenterology 1979;77:1091-3.
54. Rump JA, Scholmerich J, Gross V, et al. A new type of perinuclear antineutrophil cytoplasmic antibody (p-ANCA) in active ulcerative colitis but not Crohn's disease. Immunobiology 1990;181:406-13.
55. Saxon A, Shanahan F, Landers CJ, et al. A distinct subset of antineutrophil cytoplasmic antibodies is associated with inflammatory bowel disease. J Allergy Clin Immunol 1990;86:202-10.
56. Lightfoot RW Jr, Michel BA, Bloch DA, et al. The American College of Rheumatology 1990 criteria for the classification of polyarteritis nodosa. Arthritis Rheum 1990;33:1088-93.
57. Beattie RM, Nicholls SW, Domizio P, et al. Endoscopic assessment of the colonic response to corticosteroids in children with ulcerative colitis. J Pediatr Gastroenterol Nutr 1996;22:373-9.
58. Brogan PA, Davies R, Gordon I, et al. Renal angiography in children with polyarteritis nodosa. Pediatr Nephrol 2002;17:277-83.
59. Jennette JC, Falk RJ, Andrassy K, et al. Nomenclature of systemic vasculitides. Proposal of an international consensus conference. Arthritis Rheum 1994;37:187-92.
60. Hunder G. The use and misuse of classification and diagnostic criteria for complex diseases. Ann Intern Med 1998;129:417-8.
61. Petty RE, Cassidy JT. Vasculitis and its classification. In: Cassidy JT, Petty RE, eds. Textbook of Pediatric Rheumatology. Philadelphia: WB Saunders Co, 2001:564-8.
62. Albert DA, Silverstein MD, Paunicka K, et al. The diagnosis of polyarteritis nodosa. II. Empirical verification of a decision analysis model. Arthritis Rheum 1988;31:1128-34.
63. Albert DA, Rimon D, Silverstein MD. The diagnosis of polyarteritis nodosa. I. A literature-based decision analysis approach. Arthritis Rheum 1988;31:1117-27.
64. Brogan PA, Dillon MJ. Vasculitis from the pediatric perspective. Curr Rheumatol Rep 2000;2:411-6.
65. Nash MC, Dillon MJ. Antineutrophil cytoplasm antibodies and vasculitis. Arch Dis Child 1997;77:261-4.
66. Kerr GS, Fleisher TA, Hallahan CW, et al. Limited prognostic value of changes in antineutrophil cytoplasmic antibody titer in patients with Wegener's granulomatosis. Arthritis Rheum 1993;36:365-71.
67. Brogan PA, Dillon MJ. The use of immunosuppressive and cytotoxic drugs in non-malignant disease. Arch Dis Child 2000;83:259-64.
68. Booth A, Harper L, Hammad T, et al. Prospective study of TNF alpha blockade with infliximab in anti-neutrophil cytoplasmic antibody-associated vasculitis. J Am Soc Nephrol 2004;15:717-21.
Keywords:

vasculitis; inflammatory bowel disease; polyarteritis nodosa; ulcerative colitis; Crohn disease

© 2006 Lippincott Williams & Wilkins, Inc.