What Is Known/What Is New
What Is Known
- Twenty-four cases of pediatric patients with diagnoses of both inflammatory bowel disease (IBD) and chronic recurrent multifocal osteomyelitis (CRMO) are previously reported.
- There is limited understanding of the pathophysiology of the association and limited data that includes newer treatments (biologic agents).
What Is New
- Our case series describes the largest single-center patient group to date.
- We report two cases of patients with IBD, CRMO, and psoriasis which has not been previously reported.
- Our report includes the first use of ustekinumab for the dual diagnosis of IBD and CRMO in children.
- The response to tumor necrosis factor (TNF)-α inhibitors may identify a possible pathophysiologic link between the gut–synovial axis and elevated TNF-α levels seen in patients with CRMO.
Up to 50% of pediatric patients with inflammatory bowel disease (IBD) will develop at least one extraintestinal manifestation in their lifetime (1). Musculoskeletal involvement is the most common extraintestinal manifestation, including inflammatory peripheral (oligoarticular asymmetric or polyarticular symmetric) and axial arthritis (sacroiliitis, inflammatory back pain, and ankylosing spondylitis), as well as non-inflammatory features including osteopenia, osteonecrosis and compression fractures (2). The first reported case of a dual diagnosis of IBD and chronic recurrent multiple osteomyelitis (CRMO) was by Bognar et al (3), although a previous review of CRMO had noted 5 of 30 patients also had a diagnosis of IBD (4). Since then, approximately 32 cases have reported the association, 24 of which are pediatric (5–8). Given the relative paucity of literature exploring this association, there remains a lack of understanding of the etiology, course, and prognosis in pediatric patients. To date, no common genetic marker has been identified in patients with both diagnoses. A mutation of FBLIM1 has been recently described in two patients with CRMO, as well as demonstrated in mice with CRMO (9), however further research is needed to delineate if there is any genetic association between IBD and CRMO.
The McMaster Children's Hospital patient database was searched for patients <18 years old with a diagnosis of IBD or CRMO from January 1, 2010 to June 30, 2020. The lists of patients with each diagnosis were compared to identify patients with a dual diagnosis of IBD and CRMO. A retrospective chart review was performed for timelines of diagnoses, patient characteristics, and treatment courses. A descriptive analysis of the data was performed.
Of 600 patients with IBD and 47 with CRMO identified, seven pediatric patients were found to have a diagnosis of both IBD and CRMO. All of these patients were followed by both pediatric gastroenterology and rheumatology (Table 1).
TABLE 1 -
Demographics and IBD features of seven patients with IBD and CRMO
||Age at IBD onset (y)
||Paris classification of disease phenotype
||Initial IBD therapy
||Age at CRMO onset (y)
||Years between diagnoses
||Latent TB EBV hepatitis Psoriasis vulgaris
||Ileal (L1) B2
||EEN and methotrexate
||Iron deficiency anemia
||Colonic (L2) B1p
||Left-sided disease (E2)
||Ileal (L1) B1
||EEN + methotrexate SC
||Pityriasis amiantacea Pityriasis alba Sebaceous nevus
Behavior: B1 = non-stricturing, non-penetrating; B2 = stricturing; CD = Crohn disease; CD Paris Classification: Location: L1 = distal third ileum with limited cecal; L2 = colonic; L3 = ileocolonic; CRMO = chronic recurrent multifocal osteomyelitis; F = female; G1 = growth delay; Growth: G0 = no growth delay; IBD = inflammatory bowel disease; IBD-U = inflammatory bowel disease-unclassified; M = male; p = perianal disease modifier; PMH = past medical history; UC = ulcerative colitis; UC Paris Classification: E2 = disease distal to the splenic flexure; E3 = extensive (from the hepatic flexure distally); E4 = pancolonic (disease extends proximal to the hepatic flexure).5-ASA = 5- aminosalicylic acid; CMPA = cow's milk protein allergy; EBV = Epstein-Barr virus; EEN = exclusive enteral nutrition; SC = subcutaneous; TB = tuberculosis.
∗Disease was only noted on histology following steroid course.
The cohort consisted of five females and two males. In contrast to previous reports (7), the majority of these patients (6/7, 86%) had an initial diagnosis of IBD and were later diagnosed with CRMO (Table 1). Of those patients, three had Crohn disease (CD), two had ulcerative colitis (UC), and one was labeled as IBD unclassified (IBD-U).
At the time of their second diagnosis, patients ranged in age from 3 to 16 years old. The median (interquartile range [IQR]) age at IBD diagnosis was 9.0 (2.5) years, while at CRMO diagnosis was 12.5 (4.0) years. The median time between the two diagnoses was 2.3 (2.0) years (range 3 months to 5 years).
All patients with IBD who developed CRMO presented with bony pain and had findings in keeping with CRMO on diagnostic MRI (Table 2). Identified lesions were localized to the femoral metaphysis (3/7; two proximal, one distal), proximal tibia (2/7), clavicle (2/7), and mandible (1/7).
TABLE 2 -
Presentation of seven patients with CRMO
||IBD therapy at CRMO onset
||Current CRMO management
||Right clavicle pain and swelling, psoriatic plaques on hands, knees, scalp and left foot
||Elevated CRP (20.7 mg/L) ANA, RF, and HLA-B27 negative
||Solitary, non-specific, right clavicle lesion
||Left maxilla pain and swelling
||Elevated CRP (9.6 mg/L)
||Left mandibular patchy marrow edema with associated periosteal reaction
||Adalimumab q2wks + sulfasalazine
||Right 3rd and left 5th toe pain and left medial clavicle pain and swelling
||Normal CRP (0.8 mg/L)ANA positive (1:640)
||Right knee pain, weight loss
||Elevated CRP (13.9 mg/L) ANA and ANCA negativeHLA-B27 inconclusive
||Bilateral distal femoral metaphysis, bilateral proximal tibial metaphysis, bilateral posterior aspects of the calcaneus and left first metatarsal bone marrow edema
||Adalimumab q1wk + methotrexate SC
||Left hip, thigh and ankle pain; right thigh pain
||Elevated CRP (38.7 mg/L)HLA-B27 negative
||Focal areas of abnormal enhancement affecting the right proximal femur and left acetabulum
||Adalimumab q1wk + methotrexate PO
||Left clavicle, bilateral knee and ankle pain; left clavicle swelling
||Normal CRP (1.6 mg/L)
||Findings in keeping with osteomyelitis of the medial half of the left clavicle and T8 vertebral body
||Infliximab q7wks + methotrexate SC
||Left knee, tibia and ankle pain; left medial tibial swelling
||Elevated CRP (26 mg/L)Elevated ESR (41 mm/h)
||Multiple patchy areas of hyperintense marrow edema mainly in the distal femurs and proximal tibias
ANA = antinuclear antibody; ANCA = antineutrophil cytoplasmic antibody; CRMO = chronic recurrent multifocal osteomyelitis; CRP = C-reactive protein; MRI = magnetic resonance imaging; PO = oral; qXwk(s) = every X weeks; RF = rheumatoid factor; SC = subcutaneous.
At the time of CRMO diagnosis, patients’ IBD therapy included at least one of sulfasalazine (2/6), infliximab (3/6), adalimumab (1/6), or no treatment (1/6) (Table 2). Due to the underlying diagnoses of IBD, the gastroenterology team was involved in initial management of CRMO, as the use of NSAIDs was controversial. Two patients were started on a short course of naproxen (Cases 1 and 2), both of whom have since remained stable on their anti-TNF-α therapy for IBD. Of note, Case 1 also has a diagnosis of psoriasis, which preceded his CRMO diagnosis. One patient (Case 3) was started on celecoxib for their CRMO and has remained on that treatment with no CRMO relapses.
One patient (Case 4) was not on IBD treatment at the time of diagnosis due to ongoing diagnostic evaluation. At CRMO diagnosis, he was started on subcutaneous methotrexate. After evaluation by both teams, he was also started on adalimumab biweekly, which was later increased to weekly for control of ongoing IBD-related symptoms. Both diseases are now well controlled on weekly subcutaneous methotrexate and adalimumab. Case 5 was on sulfasalazine at the time of CRMO diagnosis and was started on oral methotrexate as a result of the diagnosis. A month later, adalimumab was added for treatment of both CRMO and IBD and both conditions have been well controlled since. Case 6 was on infliximab at the time of CRMO diagnosis and no changes were made to her management. Eventually, she developed psoriasis, thought to be induced by infliximab, and was switched to ustekinumab (an interleukin [IL]-12/IL-23 inhibitor), which is controlling all conditions. To the best of our knowledge, this is the first documented case of IBD and CRMO treatment with ustekinumab.
The one patient (Case 7) diagnosed with CRMO before diagnosis had an elevated fecal calprotectin upon screening investigations, performed due to physician awareness of the co-existence of both conditions. She was subsequently referred to gastroenterology and underwent a colonoscopy and endoscopy which were grossly normal; however, biopsies revealed mild, chronic and active colitis of the transverse and sigmoid colon. She was diagnosed with IBD-U and started on 5-aminosalicylic acid (5-ASA). At the time of diagnosis, her CRMO management included naproxen, although she continued to experience significant bony pain. Due to these ongoing symptoms, she was trialed on sulfasalazine and after limited response, as well as new development of abdominal pain, she was started on infliximab every 4 weeks. Both conditions are now well controlled on this therapy.
We have identified that 1% of pediatric patients with IBD at our center also had CRMO. Although the etiology of the link remains unknown, most patients had an initial diagnosis of IBD and were later diagnosed with CRMO. This contrasts with previous reports (7) where the CRMO diagnosis often preceded the IBD diagnosis. This difference may be explained by improved recognition of CRMO by our gastroenterology teams at our institution. Alternatively, the identification of CRMO either before or after the diagnosis of IBD may be similar to the timing of other extraintestinal manifestations of IBD, which can also precede or follow diagnosis. The average time between the two diagnoses in our case series and the literature is 27 and 19 months, respectively (7). A slight predominance of female patients was described. Although IBD demographics are typically equivocal or slightly male predominant, this skew towards female sex is in keeping with the prevalence of female patients with CRMO (10). Both our series and previous literature report a slight predominance of CD in patients who develop both conditions, however, this is likely a reflection of the overall predominance of CD in the pediatric IBD population (10). We note that within our population, most patients had colonic disease, although two patients had only isolated small bowel (ileal) CD.
The typical locations of CRMO development in pediatric patients are lower extremity long bones, vertebrae, clavicles, and mandible (11). The population we describe demonstrates typical patterns of CRMO.
Notably, 67% of children in our series were on anti-TNF therapy for their IBD at the time of developing CRMO symptoms. Anti-TNF agents are used to treat CRMO that is refractory to NSAID therapy or for patients that require frequent NSAIDs due to relapses (12). Although the pathogenesis of CRMO is not fully understood, it seems unlikely that anti-TNF-α therapy would promote its development. In fact, one of the leading hypotheses about the pathogenesis of CRMO is the imbalance between pro-inflammatory cytokines (eg, IL-6, IL-1ß, and TNF-α) and anti-inflammatory cytokines (eg, IL-10), which play an important role in bone resorption and remodeling (13). Peripheral blood cells from patients with active CRMO stimulated in vitro with lipopolysaccharide have shown a significant increase in IL-1ß release compared to healthy controls. Similar studies on monocytes from CRMO patients have demonstrated reduced IL-10 and IL-19 and enhanced IL-20 expression, promoting inflammation in general. More specifically, and relevant to our population, elevated levels of TNF-α have been measured in up to 66% of patients with CRMO flares (14). This is the proposed mechanism of anti-TNF-α therapy in CRMO—modulating TNF-α balancing bone turnover and eliminating and preventing lesions (13).
The link between IBD and other inflammatory, musculoskeletal manifestations has been well described through the gut-synovial axis (15). Cellular and humoral immune responses access systemic circulation from the gut and gain access to joint spaces as a result. Although many inflammatory mediators are involved, TNF-α IL-12 and IL-23 are specifically important and upregulated (16,17). Some reports have described CRMO as an extraintestinal manifestation of IBD (6), which would extend the gut–synovial axis hypothesis to include bone marrow inflammation, the hallmark feature of CRMO. Upregulation of TNF-α has been modelled in mice as a contributor to the development of both IBD and arthritis (16). Further models of rats with colitis have shown increased TNF-α at the intestinal mucosa, followed by increased plasma levels of TNF-α (18). As mentioned, TNF-α has been measured in CRMO patients during disease flare and shown to be elevated in the majority of patients. This would further explain the positive response to TNF-α inhibitors in our population. Further, IL-12 has been measured and monitored in patients with CRMO, and noted as a marker of treatment response (13). Similar to TNF-α, rat models have demonstrated elevated levels of IL-23 in the plasma of rats with colitis and spondylarthritis, later demonstrated to originate from the intestinal mucosa (18). Levels of IL-23 have not been measured in patients with CRMO. We describe one patient with a response to ustekinumab, an IL-12/IL-23 inhibitor, making this therapy and the role of IL-12 and 23 of interest for further investigation in this patient population. Further research is needed to investigate these hypotheses.
Most previous reports of pediatric patients with IBD and CRMO were published before the availability or widespread use of biologic agents (5). As a result, most data describes the use of corticosteroids, 5-ASA, and azathioprine as a treatment for IBD and NSAIDs and corticosteroids for CRMO. Although a small sample size, all seven of our described patients were treated with a biologic agent, mainly TNF-α inhibitors. This information helps to support the further use of these agents in these patients. Further, we describe for the first time, the use of ustekinumab, as a therapeutic option in patients that do not tolerate TNF-α inhibitors.
Interestingly, we describe two patients that also have a diagnosis of psoriasis. Both IBD and CRMO have been previously described in patients who also have psoriasis (19). To the best of our knowledge, this triad of conditions has not been described in pediatrics. Further, there have been some descriptions of a paradoxical relationship between the initiation of TNF-α inhibitors in patients with IBD and a new onset of psoriasis, which was seen in Case 6 (20).
Our series describes the largest single-center cohort of pediatric patients with a dual diagnosis of IBD and CRMO. In patients with IBD who have prolonged episodes of bony pain, it is important to consider x-rays, or localized or whole-body MR imaging. If patients with CRMO develop chronic abdominal pain, diarrhea, or bloody stools not otherwise explained, gastroenterology consultation and further workup for IBD should be considered. Biologic agents, mainly TNF-α inhibitors (adalimumab and infliximab) are often an effective treatment option for both conditions and can reduce the burden of polypharmacy. Ustekinumab may be a reasonable treatment option for those that do not respond to or tolerate TNF-α inhibitors. In well controlled IBD, CRMO flares may be controlled with NSAIDs if safe or pamidronate, rather than changing biologic therapy. Ongoing descriptive reports and analysis are needed to continue to delineate this association.
M.D., K.B., M.S., and M.B. designed the study. M.D. acquired and interpreted data and drafted the manuscript. K.B., M.S., and M.B. interpreted the data and contributed to drafting the manuscript. All other authors read and approved the final manuscript.
1. Jang HJ, Kang B, Choe BH. The difference in extraintestinal manifestations of inflammatory bowel disease for children and adults. Transl Pediatr
2. Jose FA, Garnett EA, Vittinghoff E, et al. Development of extraintestinal manifestations in pediatric patients with inflammatory bowel disease. Inflamm Bowel Dis
3. Bognar M, Blake W, Agudelo C. Chronic recurrent multifocal osteomyelitis associated with Crohn's disease. Am J Med Sci
4. Kahn MF. Chronic recurrent multifocal osteomyelitis. Association with vertebra plana. J Bone Jt Surg
5. Audu GK, Nikaki K, Crespi D, et al. Chronic recurrent multifocal osteomyelitis and inflammatory bowel disease. J Pediatr Gastroenterol Nutr
6. Ahmed ABM, Alsaleem BMR. Unusual manifestation of ulcerative colitis
. Case Rep Pediatr
7. Van Ommen C, Dehoorne J, De Baets F, et al. A case of chronic recurrent multifocal osteomyelitis associated with Crohn's disease. Acta Gastroenterol Belg
8. Bousvaros A, Marcon M, Treem W, et al. Chronic recurrent multifocal osteomyelitis associated with chronic inflammatory bowel disease in children. Dig Dis Sci
9. Cox AJ, Darbro BW, Laxer RM, et al. Recessive coding and regulatory mutations in FBLIM1 underlie the pathogenesis of chronic recurrent multifocal osteomyelitis (CRMO). PLoS One
2017; 12:e0169687doi: 10.1371/journal.pone.0169687.
10. Ye Y, Manne S, Treem WR, et al. Prevalence of inflammatory bowel disease in pediatric and adult populations: recent estimates from large national databases in the United States, 2007–2016. Inflamm Bowel Dis
11. Twilt M, Laxer RM. Clinical care of children with sterile bone inflammation. Curr Opin Rheumatol
12. Zhao Y, Wu EY, Oliver MS, et al. Consensus treatment plans for chronic nonbacterial osteomyelitis refractory to nonsteroidal antiinflammatory drugs and/or with active spinal lesions. Arthritis Care Res (Hoboken)
13. Taddio A, Zennaro F, Pastore S, et al. An update on the pathogenesis and treatment of chronic recurrent multifocal osteomyelitis in children. Pediatr Drugs
14. Jansson A, Renner ED, Ramser J, et al. Classification of non-bacterial osteitis: retrospective study of clinical, immunological and genetic aspects in 89 patients. Rheumatology
15. Brakenhoff LKPM, van der Heijde DM, Hommes DW, et al. The joint-gut axis in inflammatory bowel diseases. J Crohns Colitis
16. Kontoyiannis D, Pasparakis M, Pizarro TT, et al. Impaired on/off regulation of TNF biosynthesis in mice lacking TNF AU-rich elements. Immunity
17. Lees CW, Barrett JC, Parkes M, et al. New IBD genetics: common pathways with other diseases. Gut
18. Hedin CRH, Vavricka SR, Stagg AJ, et al. The pathogenesis of extraintestinal manifestations: implications for IBD research, diagnosis, and therapy. J Crohns Colitis
19. Weng X, Liu L, Barcellos LF, et al. Clustering of inflammatory bowel disease with immune mediated diseases among members of a northern California-managed care organization. Am J Gastroenterol
20. Fiorino G, Allez M, Malesci A, et al. Anti TNF-α induced psoriasis in patients with inflammatory bowel disease. Aliment Pharmacol Ther