Secondary Logo

Journal Logo

Short Communications: Gastroenterology

Severe Abdominal Manifestations in Juvenile Dermatomyositis

Besnard, Caroline; Gitiaux, Cyril†,‡; Girard, Muriel†,§,||,¶; Galmiche-Rolland, Louise#; Talbotec, Cécile§; Quartier, Pierre∗,†,∗∗; Bodemer, Christine†,††; Berteloot, Laureline‡‡; Bader-Meunier, Brigitte∗,∗∗

Author Information
Journal of Pediatric Gastroenterology and Nutrition: February 2020 - Volume 70 - Issue 2 - p 247-251
doi: 10.1097/MPG.0000000000002575
  • Free


What Is Known/What Is New

What Is Known

  • Juvenile dermatomyositis may present with gastrointestinal complications in 22% to 37% of cases.
  • Reported gastrointestinal manifestations mainly involve the digestive tract.

What Is New

  • Pancreatitis should be considered as a main differential diagnosis in these patients.
  • Severe gastrointestinal manifestations were mainly caused by primary juvenile dermatomyositis involvement.
  • Severe gastrointestinal manifestations are associated with a high mortality rate.
  • Urgent evaluation and early aggressive treatment are warranted for children with active juvenile dermatomyositis who develop severe and/or sustained abdominal pain.

Juvenile dermatomyositis (JDM) is a systemic immune-mediated disease, characterized by inflammation of the skeletal muscles and skin. It represents up to 85% of pediatric idiopathic inflammatory myopathies (1). Primarily affecting skin and skeletal muscle, it may sometimes present with acute complications, especially involving the digestive tract (2,3). Such digestive tract complications occur in 22% to 37% of JDM patients (1), consisting of dysphagia, bowel dysmotility, malabsorption, and ulcerations and perforations resulting from bowel vasculitis and vasculopathy. Data, however, on the causes and course of JDM-related gastrointestinal (GI) involvement, as a whole, are still limited. Here, we conducted a retrospective single-center study to better understand the specific causes and outcomes, highlighting the need for early appropriate management of severe abdominal manifestations in JDM.


This retrospective study included patients with JDM associated with severe GI manifestation. Subjects were eligible for inclusion if they met definite criteria for JDM according to the conventional clinical-pathological criteria (4), had severe involvement of any intraabdominal organ, either at diagnosis or within the course of JDM, and were followed in the referral center for rare pediatric rheumatism and systemic autoimmune diseases (RAISE) at Necker Hospital from January 2005 to December 2018. Severe GI manifestation was defined as potential life-threatening abdominal organ involvement and/or a need for abdominal surgery and/or parenteral nutrition. We recorded the involvement of all intraabdominal organs: liver, spleen, kidney, pancreas, adrenals, and digestive tract. Involvement of an abdominal organ was defined by any significant clinical, biochemical, and/or morphological abnormality. Digestive-tract involvement was defined as severe abdominal pain, dysphagia, digestive hemorrhage, obstruction, and ulceration and/or perforation. Hepatitis was defined by elevated liver enzyme levels exceeding 5 times the upper limit associated with elevated gamma-glutamyl transferase (GGT) activity and/or liver failure. According to French recommendations, the diagnosis of acute pancreatitis was considered to be certain when abdominal pain was associated with elevated blood lipase levels exceeding 3 times the upper limit. Abdominal imaging was performed in case of doubt. Muscle weakness was assessed using the Childhood Myositis Assessment Scale (CMAS) (range 0--52) and Manual Muscle Testing (MMT) (range, 0--80). Severe muscle involvement was defined by a CMAS score <15 and/or a MMT score <30. Blood was screened for the myositis-specific antibodies (MSA), anti-Mi2, MDA-5 (melanoma differentiation-associated gene 5), NXP2 (nuclear matrix protein 2), Tif1- γ (transcriptional intermediary factor-1-gamma) and the myositis-associated antibodies (MAA), anti-Ro52, Ro60, Ku, PM-Scl (polymyositis-scleroderma), RNP (ribonucleoprotein), and SS-B (Sjögren syndrome B) (5), in the same expert immunology laboratory. Deltoid muscle biopsy sampling, histological staining, and immunohistochemistry were carried out as previously described (6). All muscle biopsies were scored blindly for clinical data by C.G. using the scoring tool for severity with a validated visual analogical score (VAS) (6). Clinically inactive disease was defined by the PRINTO criteria (7). Severe GI manifestation was classified as certain or probable primary JDM involvement or a therapy-related or infection-related event. Primary JDM involvement was considered in cases of favorable outcome under specific treatment and/or the exclusion of infection or drug-induced toxicity. A treatment-related event was suspected if it occurred after the patient received medication with potential side effects, in the absence of any other cause. An infection-related event was considered if there was evidence of systemic and/or GI infection. We reviewed the medical charts with respect to demographics, clinical presentation, biological, radiological, endoscopic, and pathological findings and therapy. Statistical data are expressed as medians [minimal–maximal].

Ethical Approval

Patients were aware of the study and registered, after parental consent, in the national CEMARA (Center of Rare Diseases) database approved by the French National Committee on Informatics and Liberty.


Of the 110 JDM patients followed during the considered period, 9 (8.3%) presented with 19 different severe GI manifestations (Table 1). The median age at diagnosis of JDM was 10.7 years (range 3.7–14.4). Patients presented with skin ulcers (n = 1), cutaneous edema (n = 5), Gottron papules (n = 5), or malar rash (n = 9). The median CMAS score was 7 (range 2–48], median MMT score 47 (range 38–76), and median creatine kinase level 3812 UI/L [range 170–14,054]. Seven of 8 patients tested positive for MSA: anti-NXP2 (n = 4), anti-TIF1-γ (n = 2), and anti-MDA5 (n = 1). Seven patients underwent muscle biopsy, which all revealed a high VAS score of at least 7/10, consistent with severe myofiber and vascular lesions. Median follow-up was 3 years (range 0.4–6.3). The first line of JDM treatment consisted of prednisone alone, until 2012, and a combination of prednisone with methotrexate thereafter. Eight patients did not respond to first-line treatment and received 2 to 7 lines of treatment before the occurrence of abdominal complications: intravenous immunoglobulins, rituximab, plasma exchanges, mycophenolic acid, hydroxychloroquine, ciclosporin A, cyclophosphamide, and ruxolitinib.

Causes, treatment, and outcome of each abdominal manifestation

GI manifestations consisted of acute pancreatitis (n = 4), digestive tract involvement (n = 11), and hepatitis (n = 4). These manifestations occurred either at the diagnosis of JDM (n = 6), or during follow-up (time interval from diagnosis: 16.5 months [range 3–27]) (n = 13). All complications, except 2 cases of hepatitis, were associated with clinically active JDM. The revealing symptom for digestive-tract involvement and acute pancreatitis was abdominal pain.

Pancreatitis was found in 4 patients, all who had elevated blood lipase levels and suggestive imaging findings by CT or MRI, namely an edematous and heterogeneous pancreas. Patient 6, with refractory JDM, had pancreatic necrosis, which did not respond to JDM-specific treatment or surgery and evolved towards chronic necrotizing pancreatitis (Fig. 1A). The pancreatitis resolved in the remaining 3 patients following JDM-specific treatment with corticosteroids and rituximab, associated with plasma exchange and exclusive parenteral nutrition. Gallbladder disease or involvement of the biliary ducts were excluded by abdominal imaging. No evidence of infection or treatment side effects were retrieved.

Radiological and histological features of the pancreas, liver, and gastrointestinal involvement of patient 6. Abdominal imaging. (A) CT scan, axial section, portal phase: (a) heterogeneous hepatomegaly, (b) edematous hypertrophic pancreas, (c) fluid collection. (B and C) MRI, axial section, T2 with fat saturation, 1 month later: (a) steatotic hepatomegaly, (b) irregular atrophy of the pancreas with ectasia of the pancreatic duct, (c) fluid collection, (d) colon wall thickening, (e) fibrofatty proliferation. 2. Liver biopsy.

Digestive tract involvement presented in 7 patients as at least 1 occurrence of severe and/or persistent abdominal pain, vomiting, and/or severe obstruction (Table 1). Abdominal imaging (CT scan and/or MRI) performed on all 7 patients resulted in normal images (n = 2), colon (n = 2) and/or ileum (n = 2) wall thickening, or bowel ulcerations and/or perforations (n = 4) (Fig. 1B). In addition, abdominal MRI identified mesenteric fibrofatty proliferation potentially resulting from inflammatory involvement (Fig. 1C). Bowel perforation occurred in 4 patients. Digestive-tract involvement was successfully treated with JDM-specific medication (corticosteroids [n = 10], ciclosporin A [n = 1], intravenous immunoglobulins [n = 6], methotrexate [n = 5], rituximab [n = 4], mycophenolic acid [n = 2], and plasma exchange [n = 5]). Two patients underwent surgery: jejunal resection for jejunal perforation and colostomy for cecal perforation. Patient 6 developed pancolitis concomitant with severe pancreatitis during a disease flare and died from refractory JDM. No evidence of infection or drug toxicity was retrieved.

Hepatitis occurred in 4 patients, with hepatomegaly and/or abdominal pain and elevated transaminase levels and GGT activities. No autoantibodies against liver were detected and the results of serological tests for hepatitis A, B, C, and E virus, cytomegalovirus, and Epstein-Barr virus were negative. Liver function and serum bilirubin and gamma globulin levels were normal. Blood cultures yielded no bacterial growth. Abdominal imaging showed heterogeneous hepatomegaly with large hypodense (CT) and/or low-signal areas (on fat-suppressed or phase-out MRI sequences) (Fig. 1B). Two patients underwent liver biopsy, which showed massive microvacuolar steatosis without zonation, either diffuse (patient 6) or present in 1/3 of hepatocytes (patient 3) (Fig. 2A). Thin centrolobular perisinusoidal fibrosis was noted in these 2 patients (Fig. 2B). Scarce lobular inflammation was focally observed, associating polymorphonuclear neutrophils with rare lymphocytes for patient 6 and predominantly macrophages for patient 3 (Fig. 2C). Neither inflammation nor biliary duct alterations were noted within the portal tracts. Liver architecture was preserved and there was no histological argument for autoimmune hepatitis. Along with steatosis, patient 6 showed lipodystrophy with hypertriglyceridemia and impaired glucose tolerance. No specific treatment was intended for patient 3 and the JDM flare of patient 6 was refractory to all medications used. Transaminase levels and GGT activities returned to normal values under JDM treatment for patient 9. Patient 4 had methotrexate-related hepatitis.

(A) Masson trichrome staining, 5×: (a) extensive steatosis. (B) Sirius red staining, 10×: (a) steatosis and (b) perisinusoidal fibrosis. (C) Hematoxylin-Eosin coloration, 20×: (a) steatosis and (b) lymphocytes.

All pancreatitis and digestive-tract involvements were classified as a certain JDM-related event. A duodenal perforation occurring 3 days after the administration of high doses of corticosteroids (patient 1) was also considered as a possible treatment-related manifestation. Hepatitis was considered to be a JDM-related complication for 3 patients, including the 2 with massive steatosis. A fourth patient had treatment-related hepatitis (Table 1).

At the last visit, JDM was clinically inactive, with complete remission of the GI complication for 4 patients. Three patients died from refractory JDM 2.9 years (range 2–3.6) after the JDM diagnosis.


This retrospective study shows that severe GI manifestations were mainly related to primary involvement of JDM and required urgent management. Primary JDM-related severe GI manifestations consisted mostly of acute pancreatitis and digestive-tract involvement revealed by abdominal pain associated with severe muscle weakness. The identification of specific MSAs has allowed the characterization of individual subgroups of children with specific phenotypes of JDM (5). Recently, Aouizerate et al.(5) reported that marked muscle ischemia and the presence of anti-NXP2 autoantibodies are associated with more severe forms of JDM, especially with more GI involvement. Most of our patients were positive for MSA antibodies, mainly against NXP2 and TIF1-γ. As these 2 antibodies are the most frequently found MSAs in JDM (8), further studies are warranted to assess a possible link between the occurrence of severe GI manifestations and anti-TIF1-γ antibodies.

Nearly half of our patients had acute pancreatitis. Pancreatitis has been previously described in only 5 reports (9) and these data emphasize that serum lipase levels should be promptly assessed in severe JDM patients presenting with abdominal pain. Abdominal CT scan showed an edematous and heterogeneous pancreas, without features suggestive of autoimmune pancreatitis. The underlying pathophysiology is still unknown, but vasculopathy could play a role, as in severe GI involvement.

JDM-related GI vasculopathy is a known factor of severity, often leading to digestive-tract ulcerations and perforations (2,10–12). They are related to chronic noninflammatory vasculopathy of small and medium-sized arteries, leading to ischemia (11). Such vasculopathy is prominent in severe JDM and is associated with severe muscle weakness at disease onset, a low remission rate at 12 months, frequent GI involvement, and higher myopathological scores (6), as observed in our patients. High doses of corticosteroids may also contribute to the development of GI complications, as suggested by the development of digestive ulcerations and perforations shortly after pulses of methylprednisolone in JDM patients with preexisting abdominal pain and severe muscle weakness reported in the present cohort and the literature (2,11,12).

JDM-related moderate steatosis has rarely been reported, especially in association with lipodystrophy, which may develop in some JDM patients with a severe and prolonged clinical course (13,14). Here, we report, for the first time, unusual massive fatty liver infiltration in 2 JDM patients. Intravenous corticosteroid treatment, parenteral nutritional support, severe pancreatitis, and lipodystrophy may have contributed to such liver involvement.

JDM-related severe GI manifestations remitted under various specific JDM treatments in all but 3 patients, who died from refractory JDM. Over the past decades, survival among patients with JDM has increased from 67% to 99% (15). 4/101 patients followed in our tertiary center, however, had a fatal outcome, including 3 of the present series who died from a refractory primary JDM complication. During the considered period, a fourth patient without abdominal involvement died from infection. Thus, this study suggests that GI manifestations, as a whole, are also a hallmark of a severe prognosis in JDM patients, consistent with our previous results showing that severe JDM cases present more frequently with GI manifestations (5,6). The severity of both GI and muscle involvement could be related to vasculopathy, which is prominent in severe JDM (6,11).

In conclusion, urgent evaluation and early aggressive treatment in an expert center are warranted in children with active JDM who develop severe and/or sustained abdominal pain. GI involvement and pancreatitis should be considered as the main diagnoses for these patients.


1. Feldman BM, Rider LG, Reed AM, et al. Juvenile dermatomyositis and other idiopathic inflammatory myopathies of childhood. The Lancet 2008; 371:2201–2212.
2. Besançon A, Brochard K, Dupic L, et al. Presentations and outcomes of juvenile dermatomyositis patients admitted to intensive care units. Rheumatology 2017; 56:1814–1816.
3. Robinson AB, Hoeltzel MF, Wahezi DM, et al. Clinical characteristics of children with juvenile dermatomyositis: the Childhood Arthritis and Rheumatology Research Alliance Registry: Investigating Juvenile DM Through a National Multicenter Registry. Arthritis Care Res 2014; 66:404–410.
4. Brown VE, Pilkington CA, Feldman BM, et al. An international consensus survey of the diagnostic criteria for juvenile dermatomyositis (JDM). Rheumatology 2006; 45:990–993.
5. Aouizerate J, De Antonio M, Bader-Meunier B, et al. Muscle ischaemia associated with NXP2 autoantibodies: a severe subtype of juvenile dermatomyositis. Rheumatology (Oxford) 2018; 57:873–887.
6. Gitiaux C, De Antonio M, Aouizerate J, et al. Vasculopathy-related clinical and pathological features are associated with severe juvenile dermatomyositis. Rheumatology (Oxford) 2016; 55:470–479.
7. Lazarevic D, Pistorio A, Palmisani E, et al. The PRINTO criteria for clinically inactive disease in juvenile dermatomyositis. Ann Rheum Dis 2013; 72:686–693.
8. Fujimoto M, Watanabe R, Ishitsuka Y, et al. Recent advances in dermatomyositis-specific autoantibodies. Curr Opin Rheumatol 2016; 28:636–644.
9. Guarella M, Jurquet AL, Retornaz K, et al. Dermatomyosite juvénile et nouveaux auto-anticorps: à propos de quatre observations. Arch Pediatr 2015; 22:1263–1267.
10. Enders FB, Bader-Meunier B, Baildam E, et al. Consensus-based recommendations for the management of juvenile dermatomyositis. Ann Rheum Dis 2017; 76:329–330.
11. Mamyrova G, Kleiner D, James-Newton L, et al. Late-onset gastrointestinal pain in juvenile dermatomyositis as a manifestation of ischemic ulceration from chronic endarteropathy. Arthritis Rheum 2007; 57:881–884.
12. Wang I, Hsu W, Shun C, et al. Juvenile dermatomyositis complicated with vasculitis and duodenal perforation. J Formos Med Assoc 2001; 100:844–846.
13. Bingham A, Mamyrova G, Rother KI, et al. Predictors of acquired lipodystrophy in juvenile-onset dermatomyositis and a gradient of severity. Medicine 2008; 87:70–86.
14. Russo RAG, Katsicas MM, Dàvila M, et al. Cholestasis in juvenile dermatomyositis: report of three cases. Arthritis Rheum 2001; 44:1139–1142.
15. Ravelli A, Trail L, Ferrari C, et al. Long-term outcome and prognostic factors of juvenile dermatomyositis: a multinational, multicenter study of 490 patients. Arthritis Care Res 2010; 62:63–72.

abdominal pain; digestive vasculopathy; pancreatitis

Copyright © 2020 by European Society for Pediatric Gastroenterology, Hepatology, and Nutrition and North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition