Chronic Recurrent Multifocal Osteomyelitis: What is it and how Should it be Treated?

Hermann J Girschick; Christiane Zimmer; Guenter Klaus; Kassa Darge; Anke Dick; Henner Morbach


Nat Clin Pract Rheumatol. 2007;3(12):733-738. 

In This Article

Discussion of Diagnosis

CNO and, therefore, CRMO are diagnoses of exclusion ( Box 1 ). The patient had repeatedly complained of hip pain; therefore, the differential diagnosis initially included juvenile idiopathic arthritis.[1] Clinical examination, however, provided no evidence for the presence of arthritis.

The laboratory parameters are neither consistent nor predictive in CNO. For example, skin-related disorders, such as palmoplantar pustulosis, psoriasis and acne conglobata, occur in up to 30% of pediatric and adolescent patients with CNO.[2] Radiography of bone lesions in early CNO often does not reveal characteristic changes although, later in the course of the disease, osteoplastic and sclerotic changes of long bones in the extremities and clavicle can be suggestive of CNO. Vertebral compression is also a rather late radiographic sign, but further differential diagnoses of malignancy and osteoporosis must be considered on presentation of this finding. The diagnosis of CNO must, therefore, rely on the clinical picture, in addition to imaging studies.

T2-weighted MRI sequences with fat-suppression techniques (e.g. turbo inversion recovery magnitude and short-inversion-time inversion recovery) are useful at the initial and follow-up examinations. In addition, enhancement with gadolinium-diethylene triamine penta-acetic acid provides further insights into the inflammatory activity of the lesions.[3] Technetium bone scans are helpful in the initial diagnostic setting because clinically silent CNO lesions are often present. Total-body MRI and technetium bone scans seem to be equally useful for defining the overall presence of lesions. MRI is somewhat more sensitive for showing bone marrow edema; however, whole-body MRI is currently unavailable in most institutions.

In a considerable number of patients, diagnostic imaging alone does not rule out malignancy; therefore, biopsy should be considered, especially because it is often difficult to make a definite distinction between oncologic bone lesions and lesions associated with CNO.[3,4] Osteosarcoma, Ewing's sarcoma, neuroblastoma, rhabdomyosarcoma, leukemia, Langerhans' cell histiocytosis, osteoid osteoma and osteoblastoma should all be regarded as differential diagnoses. Although most of these diagnoses can usually be ruled out by the histological work-up, osteoid osteoma and osteoblastoma can have significant inflammatory components; therefore, a biopsy that is not representative of the whole lesion might be misleading. Imaging techniques, such as technetium bone scans and total-body MRI, are helpful in selecting a suitable site for biopsy. In choosing the biopsy location, functional and cosmetic aspects must be considered. Biopsies should only be performed for diagnostic purposes and clinicians should not aim to excise the whole lesion; this could lead to unnecessary functional impairment and scarring.[5]

The need for a diagnostic biopsy has repeatedly been questioned in the management of CNO. A diagnosis of CNO seems quite probable if the bone lesions have been present for 6 months or longer and the patient also presents with typical skin lesions. In this case, a biopsy might be omissible; however, a short-term clinical follow-up, including repetition of imaging studies, is mandatory. Unifocal lesions, which have a solely osteolytic appearance and involve the surrounding tissue structures, must be biopsied to exclude malignancy.[5]

The decision to take a biopsy should be carefully considered for every patient. In addition to revealing malignancy, a biopsy can reveal subperiosteal bone formation, which is a sign of chronic inflammation. Granulocytes might be apparent in early lesions, followed predominantly by lymphocytes, plasma cells and monocytes. In later stages, fibrotic changes and hyperostosis are present. A thorough histological work-up of the whole biopsy is recommended to document these features, which might be present in different areas of a single biopsy.

Although early reports considered Propionibacterium acnes to be a relevant pathogen in CNO, later studies have not confirmed this. In addition, Bartonella henselae rarely causes multifocal osteolytic inflammatory lesions. Treatment strategies involving antibiotics have not proven effective in the long term, and, in several studies, no consistent infectious agent could be detected in lesions in pediatric patients using 'state-of-the-art' microbial techniques.[2,5,6,7,8] CNO cannot be distinguished from acute or subacute bacterial osteomyelitis using a histological examination alone.[5] An extensive microbial work-up of the tissue biopsy (long-term cultures), including polymerase chain reaction that targets eubacteria and mycobacteria, is thus essential to establish the diagnosis of sterile inflammation.[5]

Although the CRMO form of CNO has been recognized as a clinical entity for more than 30 years,[9] its origin and pathogenesis are unclear. Autoinflammatory diseases, autoimmunity, errors of metabolism and postinfectious reactive inflammation are all potential areas for the focus of research. CNO might be linked to enthesitis-related arthritis (ERA; also called 'adolescent spondyloarthropathy') and psoriatic arthritis. Several studies have identified arthritis as a concomitant feature in up to 80% of patients.[2] Diagnostic criteria for ERA might not be present at the onset of symptoms; however, ERA and psoriatic arthritis have been reported in the long-term follow-up of several cohorts of children and adolescents, despite anti-inflammatory treatment.[2,7,8,10] The degree of arthritis at the time of diagnosis and during the course of CNO might be underestimated in the literature. Of note, two patients have recently been described to have developed inflammatory bone lesions similar to CRMO while suffering from hypophosphatasia, a genetically inherited metabolic bone disease.[11]

The association of CNO with chronic inflammatory bowel disease has previously been well documented and seems to involve around 10% of CNO patients.[6,12,13] CNO and Crohn's disease are chronic auto-inflammatory syndromes. In patients with familial Crohn's disease, mutations in the caspase activation and recruitment domain 15 gene (CARD15), which encodes the nucleotide-binding oligomerization domain 2 (NOD2) protein, have been described. NOD2 is a cytoplasmic protein involved in intracellular pathogen resistance and regulation of inflammation.[14,15] Polymorphisms in the tumor necrosis factor (TNF) promoter have also been associated with a severe form of Crohn's disease.[16] Similar mutations have not yet been described in CNO; however, chronicity, the recurrence of bone and joint inflammation, and the association with psoriasis in patients using TNF antagonists[17] suggest an auto-inflammatory and rheumatologic cause.

A putative mouse model exists for CNO.[18] Genetic analysis of CNO patients initially suggested a chromosomal link to a similar region of the human genome in a limited number of patients,[19] but these findings have not yet been confirmed.

In adulthood, a similar entity to the CRMO form of CNO has been termed 'SAPHO syndrome' (consisting of synovitis, acne, pustulosis, hyperostosis and osteitis).[6] Not all of these characteristic symptoms must be present for the diagnosis, however. The principal management strategy for SAPHO syndrome does not seem to be different to that for CNO. Similar to CNO, SAPHO syndrome can have overlapping symptoms with rheumatologic conditions, including psoriasis arthritis and spondyloarthropathy.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.