Chronic recurrent multifocal osteomyelitis (CRMO) is an immune-mediated chronic inflammatory self- limiting disease with non-suppurative inflammation involving one or multiple bone foci. A bacterial etiology has not yet been proven, but the investigators hypothesize that CRMO is caused by chronic infection. Using existing knowledge of chronic infections combined with sensitive molecular techniques, the hope is to elucidate the cause of CRMO by proving or disproving bacteria as the underlying etiology.

In addition, a novel method, 18F-NaF-PET/MRI, will be investigated as to which it can improve diagnosis of CRMO. The investigators hypothesize that 18F-NaF-PET/MRI will have higher diagnostic accuracy than 99m-diphosphonate planar bone scintigraphy and SPECT/CT.

Epidemiology, clinical manifestations and prognosis Chronic recurrent multifocal osteomyelitis (CRMO) is an immune-mediated chronic inflammatory self- limiting disease with non-suppurative inflammation involving one or multiple bone foci. The disease occurs primarily in children and adolescents. The estimated prevalence is 1-5/10.000 in the European population, but since there are no studies providing precise estimates, the number of cases can be both under- and overestimated.

The mean age of CRMO onset is 10 years. There is a female preponderance with twice as many girls affected.

CRMO presents with bone pain with or without swelling and warmth. Some patients have a more acute presentation with severe pain, malaise and fever. All bones except the neurocranium can be affected. Involvement of the clavicle is classical, but the metaphyses of the long bones, the mandible and the pelvis are also frequently affected. Vertebral involvement is seen in 4 to 30% of cases in retrospective studies.

Extra-skeletal involvement can include the skin, the eyes and the gastrointestinal tract. Involvement of the skin may manifest as plantar pustulosis, psoriasis acne, pyoderma gangrenosum and Sweet Syndrome. In the gastrointestinal tract, Crohn's disease or lesions mimicking this disease have been associated with CRMO.

Previously, the prognosis of CRMO patients was considered favorable, as the disease is often self-limiting, but newer studies have revealed that physical impairment may persist in up to 50% of patients. They typically suffer from chronic pain, bone deformities and have increased fracture risk. Furthermore, there is evidence that CRMO may evolve into spondyloarthropathy with clinical and radiological sacroiliitis.

Pathophysiology The pathophysiology of CRMO is largely unknown, but is regarded as an auto-inflammatory bone disorder. Recent findings indicate that an imbalance between the pro-inflammatory cytokines Interleukine-6 (IL-6) and Tumor Necrosis Factor α (TNF-α), and the anti-inflammatory cytokine IL-10 may play a central role in the molecular pathology of CRMO. High levels of IL-6 and TNF-α have been found in serum from CRMO patients. Furthermore, it has been demonstrated that monocytes in patients with CRMO fail to produce IL-10. IL-10 is an immunomodulatory cytokine that attenuates inflammation on multiple levels.

However, in bone biopsies from CRMO patients, the lesions resemble infectious osteomyelitis histologically. Initially the predominating cell types are neutrophils, but in later disease stages monocytes, macrophages, lymphocytes and plasma cells can be detected. In the final disease stage of CRMO sclerosis and/or fibrosis can be seen. Infection with propionibacterium acnes (P. acnes) or other commensal skin bacteria (i.e. Bartonella) have been suggested to trigger CRMO. P. acnes may induce the toll-like receptors in the monocyte cell membrane, resulting in mitogen-activated protein kinase (MAP kinase) activation leading to production of pro-inflammatory cytokines. This could suggest that bacteria might be the reason for the observed imbalance in the pro-inflammatory cytokines, but the presence of bacteria in biopsies from CRMO lesions has only been demonstrated in single cases. However, in the majority of reported cases the microbiological investigations were performed solely by the use of microbiological culture and very few studies have utilized modern molecular techniques. The methods applied in these studies are more than 10 years old and the microbiological investigations are far more advanced today.

Discrepancy between bacteria found by culture swaps and modern molecular techniques have been seen in chronic wounds and chronic infection in soft tissue fillers lesions. In these chronic infections the bacteria are able to form a biofilm. The biofilm itself induces an ongoing inflammatory reaction, as it provides a protected environment for the bacteria, in which they may avoid the effects of antibiotics and host immune defence. Bacteria in biofilm grow in small clusters like planktonic bacterial cells and this complicates identification of these microorganisms in biopsy material.

Furthermore, a porcine model demonstrated chronic inflammation with formation of biofilm in bone infected with Staphylococcus aureus. S. aureus is the most frequent cause of osteomyelitis in children.

Diagnosis There are no validated diagnostic criteria and no diagnostic tests for CRMO. First of all, it is important to exclude differential diagnoses as infection (mycobacterium, septic osteomyelitis etc.), malignancies (bone tumors, metastases, leukemia or lymphoma), benign tumors (osteoid osteoma, bone cysts, fibrosis etc.) or other auto-inflammatory disorders. Biopsy is often needs to exclude an infections etiology or malignancy.

Bone scintigraphy using methylene diphosphonate or hydroxymethylene diphosphonate labeled with Technetium 99m has high sensitivity in identifying symptomatic as well as asymptomatic lesions. Monitoring disease activity during treatment or follow-up is also possible with bone scintigraphy by evaluating the degree of tracer uptake in the lesions. Furthermore, the technique can be useful in targeting biopsy to the most active lesion or identifying a more easily accessible lesion than the presenting site.

Another frequently used modality to diagnose CRMO is whole-body Magnetic Resonance Imaging (MRI). A study comparing MRI and planar bone scintigraphy found higher sensitivity of MRI in identifying CRMO lesions. Especially lesions in the spine, pelvis and femora were better visualized by MRI.

The functional imaging of bone scintigraphy and the morphological information from MRI can be combined in Positron Emission tomography (PET)/MRI. PET/MRI using 18F-fluorodeoxy-glucose (18FDG) is increasingly used clinically and in research settings in pediatric patients with malignant disease. The PET-tracer fluoride 18-labeled sodium fluoride (18F-NaF) is a positron-emitting bone-seeking tracer that provides higher resolution and better signal-to-background images, compared to 99mTc-diphosphonate bone tracers. Studies of 18F-NaF-PET in pediatric patients conclude the technique is useful for investigating the cause of bone pain in children.

There are several potential advantages of 18F-NaF-PET/ MRI. Simultaneous acquisition of PET and MRI reduces the use of anesthesia and sedation, which is often required for younger children. The effective radiation dose of 18F-NaF-PET is approximately 3 milliSievert (mSv), which corresponds to the dose from a 99mTc-diphosphonate bone scintigraphy. Whole body MRI eliminates the need for low-dose CT scan used in SPECT and therefore reduces total radiation dose to the child.

The added value of 18F-NaF-PET/MRI in CRMO patients has not yet been investigated. The hypothesis is that the technique has the potential of being a novel gold-standard method for evaluating bone pathology in children, with the superior morphological images from MRI and high resolution imaging of bone metabolism from 18F-NaF-PET.

Inclusion Criteria:

• CRMO diagnosis verified by biopsy
• Age under 18 years at diagnosis
• Malignant disease excluded
• Informed written consent

Exclusion Criteria:

• Bone biopsy not performed
• Patients not eligible for MRI (metallic foreign objects, severe claustrophobia)
• Pregnancy