Paediatric rheumatology

Chronic non-bacterial osteomyelitis: another disease associated with MEFV P.O. Avar-Aydin1, Z.B. Ozcakar1, N. Cakar1, S. Fitoz2, F. Yalcinkaya1

1Department of Paediatric Rheumatology, ABSTRACT Introduction 2Department of Paediatric Radiology, Objective. Chronic non-bacterial os- Chronic non-bacterial osteomyelitis Ankara University Faculty of Medicine, teomyelitis (CNO) is an autoinflamma- (CNO) is an autoinflammatory bone Ankara, Turkey. tory bone disease of unknown aetiolo- disease primarily of children and young Pinar Ozge Avar-Aydin, MD, MBA gy. The relationship between CNO and adults and it is characterised by sterile Zeynep Birsin Ozcakar, MD familial Mediterranean fever (FMF) inflammatory bone lesions. Its multifo- Nilgun Cakar, MD Suat Fitoz, MD is not clearly documented so far. This cal form with frequent relapses is called Fatos Yalcinkaya, MD cross-sectional study aims to evaluate chronic recurrent multifocal osteomy- Please address correspondence: the clinical and laboratory character- elitis (CRMO (1)). The aetiology of the Pinar Ozge Avar-Aydin, istics of a cohort of CNO patients with- disease is still not clear; however, some Department of Paediatric Rheumatology, in the context of its relationship with susceptibility and dysregulation Ankara University Faculty of Medicine, FMF and MEFV gene mutations. in the innate immune system have been 06000 Ankara, Turkey. Methods. Demographic and clinical shown (2-4). Persistent or relapsing E-mail: [email protected] data were extracted from electronic and remitting focal bone pain, absence Received on September 23, 2020; accepted medical records of patients with CNO. of constitutional symptoms, multifocal in revised form on November 12, 2020. The MEFV gene analysis was per- bone lesions on imaging with typical Clin Exp Rheumatol 2020; 38 (Suppl. 127): formed for all patients. lesion characteristics on typical loca- S112-S117. Results. A total number of 18 patients tions, and a personal and/or family © Copyright Clinical and with CNO with a median follow-up history of CNO-associated disorders Experimental Rheumatology 2020. of 36.50 (13.00-84.00) months were suggest a diagnosis of CNO although it included in the study. Five patients remains a diagnosis of exclusion (5, 6). Key words: chronic non-bacterial (27.8%) were found to have at least Supporting its genetic and pathogenic osteomyelitis, chronic recurrent one exon 10 mutations (four with bases, CNO has been frequently asso- multifocal osteomyelitis, familial M694V and one with M680I). Four ciated with a wide spectrum of inflam- Mediterranean fever, MEFV gene of them (22.2%) had homozygous or matory disorders, including but not compound heterozygous mutations of limited to inflammatory bowel disease the MEFV gene. Two patients had a (IBD), arthritis, and skin diseases (7). previous diagnosis of FMF and devel- Familial Mediterranean fever (FMF) is oped CNO while FMF was under con- a well-defined hereditary autoinflam- trol. Patients with MEFV mutations matory disease typically presented had an earlier onset of CNO, higher with recurrent and self-resolving epi- acute phase reactants, lower haemo- sodes of fever and serositis (8). The globin concentrations, and a higher responsible gene called the MEditer- number of bone lesions at disease on- ranean FeVer (MEFV) gene is located set with a persistent course of disease on 16p and encodes a more frequently. known as pyrin. The mutated Conclusion. Our results demonstrated pyrin causes augmented an increased frequency of MEFV gene with excessive IL-1β and NF-κβ secre- mutations in CNO and a more severe tion that induces a pro-inflammatory disease phenotype of CNO in patients condition both in patients with FMF with MEFV gene mutations. Physi- and in asymptomatic carriers of MEFV cians practicing in regions where FMF mutations (9). This pathogenetic prin- is prevalent should be aware of this re- ciple of the disease causes persistent Funding: departmental funds were used lationship and ask about the symptoms subclinical inflammation in a substan- for MEFV gene testing in patients without of FMF in detail in patients with CNO. tial portion of FMF patients, facilitates typical FMF presentation. Moreover, FMF should be included in the coexistence of other inflammatory Competing interests: none declared. CNO-associated conditions. diseases, and further, the MEFV gene

S-112 Clinical and Experimental Rheumatology 2020 CNO with MEFV mutations / P.O. Avar.Aydin et al. carriers to be associated with totally contrast sagittal spin-echo T1-weight- 232-20). Signed informed consent was different inflammatory diseases other ed (T1W), coronal gradient-echo T1W, obtained from the patients and their than FMF (9-11). Several diseases coronal short tau inversion recovery parents. such as vasculitis, juvenile idiopathic (STIR), axial plane diffusion-weighted arthritis (JIA), and IBD have been as- single-shot echo-planar imaging (b- Statistical analysis sociated with FMF but CNO is not values of 0 and 1000 s/mm2). Contrast- The IBM SPSS Statistics 21.0 software one of those (10). Moreover, although enhanced coronal and sagittal plane (IBM Corp. Released 2012. IBM SPSS bone involvement is a widening spec- T1W images with a slice thickness of Statistics for Windows, version 21.0. trum of several hereditary autoinflam- 3 to 6 mm were obtained in patients Armonk, NY: IBM Corp) was used to matory diseases, the relation of CNO with abnormal signal changes in the perform statistical analysis. Results with FMF is not clearly reported (12). pre-contrast MR series. Regional MRI were described with mean ± standard Beginning from 2010, a few case re- protocol was composed of pre-contrast deviation (SD) or median (minimum- ports suggested a link between CNO T1W three plane imaging, coronal maximum) for continuous variables and FMF/MEFV gene mutations and STIR, axial fat-saturated T2-weight- and with frequency (n) and percentage the use of colchicine as a treatment op- ed (T2W), axial diffusion-weighted (%) for categorical variables. The Kol- tion for CNO (13-16). After persistent single-shot echo-planar imaging, and mogorov-Smirnov test was used to ex- and multifocal episodes of CNO have post-contrast three plane imaging with amine the data for normal distribution. developed in two of our patients with fat saturation. Radiologic images were One-way analysis of variance (ANO- FMF, we aimed to evaluate clinical evaluated by only one paediatric radi- VA) was used for homogeneity of the characteristics, treatment modalities, ologist (SF) experienced in CNO. variables, Student’s T-test was used for and response to therapies in a cohort of To classify patients, the clinical scoring parametric data. The Mann-Whitney CNO patients within the context of its system by Jansson et al. was used that U, ANOVA, and Spearman’s correla- relationship with FMF and MEFV gene had a positive predictive value over tion were used for non-parametric data. mutations. 97% for the diagnosis of CNO if the The correlation was considered mean-

clinical CNO score was ≥39 (5). Treat- ingful if the correlation coefficient (rs) Materials and methods ment responses were evaluated accord- was >0.40. The statistical significance Patients with a diagnosis of CNO fol- ing to the clinical symptoms and imag- level was accepted as a p-value <0.05. lowed-up at the Department of Paedi- ing findings on MRI. Patients with only atric Rheumatology of Ankara Univer- one episode during the follow-up were Results sity School of Medicine were included accepted as “nonrecurrent” whereas Patient characteristics, laboratory in the study. Demographic and clinical those with relapses as “recurrent”. findings, and treatment modalities findings, results of blood tests, bone Patients who did not experience clini- A total number of 18 patients with CNO biopsies, bone marrow aspirations, and cal and radiological remission despite were included in the study. The demo- microbiological cultures, radiological therapies for 6 months were classified graphic, clinical, and laboratory char- findings, and treatment modalities were as “persistent”. Solitary lesions on acteristics of the patients with CNO are recorded from the electronic medical WBMRI were categorised as “unifo- presented in Table I. All patients had records of patients. The diagnosis of cal” and multiple lesions as “multifo- bone pain with a median diagnostic de- CNO was based on clinical symptoms cal”. Clinical remission was defined as lay time of 11.00 (2.00-48.00) months. and imaging features after the exclu- not having any symptoms for the last One patient had been treated with a sion of CNO-mimicking diseases, such 6 months and radiological remission as diagnosis of JIA three years before as infection and malignancy. Patients without any active lesions on WBMRI. the onset of CNO. Two patients had a with a possible diagnosis of CNO or Patients with a previous diagnosis of previous diagnosis of FMF and one of with a follow-up less than 6 months FMF according to Yalçınkaya criteria them also had IBD. None had a fam- were excluded. had already had the MEFV gene analy- ily history of CNO or CNO-associated Regional magnetic resonance imaging sis before the diagnosis of CNO (17). diseases whereas two patients had a (MRI), whole-body MRI (WBMRI), For others, the MEFV gene analysis family history of FMF. and bone scintigraphy, if WBMRI was performed. At least six mutations Four patients had persisting anaemia was not available, were used to assess in the MEFV gene including p.M694V, compatible with anaemia of chronic bone lesions. Whole-body MRI was p.M694I, p.M680I, p.V726A, disease. Fourteen patients had in- performed with 1.5 Tesla MR system p.E148Q, and p.K695R were analysed creased C-reactive protein (CRP) and (Achieva, Philips Healthcare, Best, by direct sequencing of the PCR-am- 10 had increased erythrocyte sedimen- The Netherlands). The coronal plane plified fragments for exon 10- muta tation rate (ESR) at the time of diagno- was the principal plane by its extensive tions and by PCR-restriction fragment sis. One patient received antibiotics be- coverage and detailed evaluation of the length polymorphism protocol for exon cause of the suspected bacterial osteo- axial skeleton. Sagittal and axial plane 2 mutations. myelitis before the diagnosis of CNO. images were added when necessary. The study was approved by the Ethics Twelve patients had bone biopsy/bone The WBMRI protocol included pre- Committee of Ankara University (#I4- marrow aspiration because most of

Clinical and Experimental Rheumatology 2020 S-113 CNO with MEFV mutations / P.O. Avar.Aydin et al.

Table I. Demographic, clinical and laboratory characteristics of patients with chronic non- No significant correlation between the bacterial osteomyelitis*. age of disease onset, laboratory param- Characteristics Patients (n=18) eters, number of bone lesions, and clin- ical CNO scores was found (rs <0.40; Female:Male 5 (27.8%):13 (72.2%) p>0.05). Age at disease onset, years 8.50 (1.50-16.00) The median follow-up was 36.50 Age at diagnosis, years 10.50 (3.50-16.17) (13.00-84.00) months. All patients re- Symptoms ceived naproxen at 15-20 mg/kg/day Bone pain 18 (100%) twice daily. Oral corticosteroids up to Point tenderness 10 (55.6%) Arthritis 6 (33.3%) 6 weeks were used in five patients as Sacroiliac joint pain 5 (27.8%) a transitional drug to other agents. A Local warmth 3 (16.7%) total of nine patients (50%) received Fever 3 (16.7%) bisphosphonate, methotrexate, tumour Acne 1 (5.6%) Bone biopsy and/or bone marrow aspiration 12 (66.7%) necrosis factor-α inhibitor (TNFi), or Normal histopathologic findings 6 (50%) a combination of these. Clinical remis- Presence of inflammation 5 (41.6%) sion was achieved in half of the study Presence of fibrosis 1 (8.3%) population (n:9) only with naproxen. Laboratory findings at disease onset The disease course was persistent in 9 count, *10 /L 10,107 ± 2,326 9 (50%) and recurrent in 6 patients count, *109/L 6,216 ± 2,082 Platelet count, *109/L 433,388 ± 168,480 (33.3%). Haemoglobin concentration, gr/dL 12.32 ± 1.68 C-reactive protein, mg/L 27.93 ± 31.63 MEFV gene analysis and comparison Erythrocyte sedimentation rate, mm/h 18.61 ± 11.55 of patients with and without ANA positivity 4 (22.2 %) HLA-B27 positivity 1 (5.6 %) MEFV gene mutations Five patients out of 18 CNO patients MRI findings at disease onset Increased signal intensity on STIR 18 (100%) (27.8%) were found to have MEFV Lesions on long bones 10 (55.6%) mutations including four (22.2%) with Metaphyseal 9 (50%) homozygous or compound heterozy- Metaphysiodiaphyseal 1 (5.6%) gous mutations. Two patients devel- Vertebral lesions 10 (55.6%) Vertebral compression 2 (11.2 %) oped CNO 2-4 years after the diag- Synovial thickening and joint effusion 8 (44.4%) nosis of FMF while FMF was under Sacroiliitis 8 (44.4%) control on a regular dose of colchicine Clavicular lesions 4 (22.2%) at 1 mg/m2/day. Patient 1 was concur- Radiographic sclerotic lesions 7 (38.9%) Radiographic osteolytic lesions 4 (22.2%) rently diagnosed with FMF and IBD Clinical CNO score 42.50 (26.00-54.00) at the age of 18 months. Although he did not have frequent FMF attacks in Treatments for CNO Naproxen 18 (100%) the first year after the diagnosis, he TNFi 7 (38.9%) experienced chronic disease anaemia Etanercept 3 (16.7%) and poor growth with continuous sub- Adalimumab 4 (22.2%) clinical inflammation and anakinra was Corticosteroids 5 (27.8%) Colchicine 5 (27.8%) used for 2 years in addition to other Methotrexate 2 (11.2%) therapies. His symptoms of CNO was Biphosphonates 2 (11.2%) started 6 months after the cessation of anakinra. Patient 2 was diagnosed ANA: anti-nuclear antibody; CNO: chronic non-bacterial osteomyelitis; FMF: familial Mediterranean fever; HLA: leukocyte antigen; MRI: magnetic resonance imaging; STIR: short tau version with CNO after complaining of persis- recovery; TNFi: tumour necrosis factor-α inhibitor. tent back and leg pain two years after *mean ± SD, median (min-max), or n (%). FMF onset. Both patients’ symptoms relieved after the first month of TNFi. them were referred to our clinic after number of 6.50 (3.00–17.00) and 15 Patient 1 experienced clinical and ra- the exclusion of malignancy. Imaging patients (83.3%) had symmetric le- diological remission at the first year of findings of all patients reported from sions. Overall, 17 patients (94.4%) TNFi. The other three patients found our centre were compatible with CNO had lower extremity involvement and to have MEFV gene mutations did not and the majority of patients (88.9%) 3 patients (16.7%) had upper extrem- have any typical clinical symptoms of had a clinical CNO score over 39 at ity involvement. None of the lesions FMF. Patient 3 was treated with TNFi disease onset. on long bones were isolated diaphyseal for 18 months and his treatment was All patients had multifocal bone le- or epiphyseal lesions. One patient with ceased after having both clinical and sions on MRI with a median lesion sacroiliitis had HLA-B27 positivity. radiological remission. However, his

S-114 Clinical and Experimental Rheumatology 2020 CNO with MEFV mutations / P.O. Avar.Aydin et al.

Table II. Characteristics of patients with chronic nonbacterial osteomyelitis and MEFV gene mutations.

Patient 1 Patient 2 Patient 3 Patient 4 Patient 5

Gender Male Male Male Male Male Age at CNO onset, years 4.5 5 10 1.5 2 Disease course of CNO Persistent multifocal Persistent multifocal Persistent multifocal Persistent multifocal Persistent multifocal Age at FMF onset, years 1.5 3 - - - Clinical findings of FMF Recurrent fever, Recurrent fever, - - - abdominal pain abdominal pain MEFV gene M694V/M694V M694V/V726A M694V/V726A M694V/- M680I/V726A Family history of FMF + - - - - Inflammatory diseases other FMF, IBD FMF Acne fulminans - - than CNO Location of inflammatory Lower extremity, Lower extremity, Lower extremity, Lower extremity, Pelvis, vertebrae bone lesions pelvis, occipital pelvis, vertebrae pelvis, vertebrae, upper extremity, condyles, vertebrae clavicula vertebrae Fever at CNO onset - - + + - Increased acute phase reactants + + + + + at CNO onset Sacroiliitis - + + - + Treatment before CNO Colchicine, oral Colchicine - - - diagnosis corticosteroid, mesalazine, azathioprine, anakinra Treatment after CNO Colchicine, naproxen, Colchicine, naproxen, Naproxen, oral Naproxen, Naproxen, diagnosis TNFi TNFi corticosteroid, biphosphonate, biphosphonate, TNFi, colchicine TNFi, colchicine TNFi, colchicine

CNO: chronic non-bacterial osteomyelitis; FMF: familial Mediterranean fever; IBD: inflammatory bowel disease; MEFV: Mediterranean FeVer; TNFi: tumour necrosis factor-α inhibitor. symptoms recurred after one month. statistical difference was not signifi- lence of FMF and the carrier rates of After his MEFV gene mutation was cant (p=0.410; OR=2.400; 95% con- MEFV gene mutations reported from documented, both TNFi and colchicine fidence interval (CI)=0.210–19.784). Turkey (18, 19). Moreover, patients were initiated concurrently, and he was All patients with MEFV mutations with MEFV mutations had an earlier followed with clinical remission for the had a persistent disease course that onset of CNO with higher acute phase last six months. After combined initia- was significantly more prevalent than reactants and a higher number of bone tion of colchicine and TNFi for patients it was observed in patients without a lesions and they showed a persistent 4 and 5, they had been followed with mutation (p=0.009; OR=0.308; 95% course more frequently suggesting clinical remission. Patient and disease CI=0.136–0.695). Tumour necrosis a more severe disease phenotype of characteristics of CNO patients with factor-α inhibitors were used signifi- CNO. MEFV gene mutations are shown in cantly more common in patients with Familial Mediterranean fever is Table II. a mutation (p=0.001; OR=0.154; 95% thought of as the best understood of the Overall, the total allelic frequency of CI=0.043–0.550). A comparison of hereditary periodic fever syndromes; MEFV genes in the study population disease characteristics of patients with however, it is a heterogeneous disease was 25%. The frequency of the M694V and without MEFV gene mutations is representing different clinical mani- allele was 13.9%. Patients with MEFV presented in Table III. festations and associated with several mutations had a significantly earlier other inflammatory conditions (10). onset of CNO, lower haemoglobin Discussion Although the disease has been increas- concentrations, higher CRP and ESR In this study, we evaluated the fre- ingly reported from all over the world, levels, and a higher number of bone le- quency of MEFV gene mutations in a it is still most prevalent in the Medi- sions at disease onset than the patients cohort of patients with CNO and found terranean and Middle Eastern regions, without a mutation. Although there was that 28% of the patients had at least one particularly in Turks, Armenians, 2.4 times increased risk of sacroiliitis and 22% had two mutations. These fre- Jews, and Arabs (20). There are many in patients with MEFV mutations, the quencies were higher than the preva- genetic and epigenetic factors affect-

Clinical and Experimental Rheumatology 2020 S-115 CNO with MEFV mutations / P.O. Avar.Aydin et al.

Table III. Comparison of disease characteristics of CNO in patients with and without caused by homozygous LPIN2 muta- MEFV gene mutations*. tions presents with early-onset CNO Patients with Patients without p-value associated with congenital dyseryth- MEFV gene MEFV gene ropoietic anaemia (31). Secondly, the mutations (n=5) mutations (n=13) deficiency of IL-1 receptor antagonist (DIRA) has a neonatal onset of CNO, Age at disease onset, years 4.60 ± 3.38 9.57 ± 3.72 0.020 periostitis, and pustulosis (32). Laboratory findings at disease onset In 2010, the first case report of a CRMO White blood cell count, *109/L 9,858 ± 1,347 10,203 ± 2,650 0.787 Platelet count, *109/L 480,600 ± 196,849 415,230 ± 161,278 0.478 patient without typical FMF symptoms Haemoglobin concentration, gr/dL 10.92 ± 1.62 12.86 ± 1.42 0.024 but carrying MEFV gene mutation was Mean corpuscular volume, fL 70.56 ± 7.05 78.26 ± 4.78 0.018 released from Japan and the patient Red-cell distribution width, % 14.54 ± 2.09 13.87 ± 1.13 0.403 benefitted from colchicine treatment C-reactive protein, mg/L 65.74 ± 38.30 13.40 ± 10.21 0.000 (13). Afterward, four other cases with Erythrocyte sedimentation rate, mm/h 27.80 ± 14.49 15.07 ± 8.39 0.032 Number of bone lesions on MRI at disease onset 11.60 ± 4.50 6.53 ± 4.05 0.035 concurrent CNO and FMF originated Sacroiliitis on MRI 3 (60%) 5 (38.5%) 0.410 from Arabic, Jewish, and Turkish de- Clinical CNO score 43.80 ± 9.44 43.76 ± 7.79 0.994 scents were reported (14-16). Although The use of TNFi 5 (100%) 2 (15.3%) 0.001 a study from Iran claimed that there Persistent course of the disease 5 (100%) 4 (30.8%) 0.009 was no significant relationship between CNO: chronic nonbacterial osteomyelitis; FMF: familial Mediterranean fever; MEFV: Mediterranean MEFV genes and CNO, recently, an in- FeVer; MRI: magnetic resonance imaging; TNFi: tumour necrosis factor-α inhibitor creased frequency of FMF among CNO * mean ± standard deviation (SD) or n (%). patients was reported (33, 34). Cicek et al. reported FMF as the most common ing phenotypic presentations. Some frequency (13.9%) and total allelic fre- concomitant disease with a frequency patients with FMF, majority carrying quency (25%) in our CNO cohort were of 34.8% in patients with CNO. They M694V mutations, had an earlier on- significantly higher in the study popu- did not find any differences in clinical set of disease, a relatively more severe lation compared to the Turkish popula- and laboratory findings of CNO- pa disease phenotype, persistent subclini- tion (4% and 8.4%, respectively (19)). tients with or without concomitant dis- cal inflammation, and a higher associa- Although the actual incidence and eases (34). On the other hand, the com- tion with other inflammatory diseases prevalence of CNO are still not known, parison of the disease characteristics of (9, 21, 22). Likewise, asymptomatic the increasing availability and tenden- CNO between patients with and with- MEFV carriers with high-risk muta- cy to perform MRI and the heightened out FMF was not defined in the study. tions showed an increase in acute phase state of awareness about the disease In our study, patients with MEFV gene reactants suggesting a proinflammatory lead more patients to be diagnosed. mutations had a significantly earlier condition among these people (9). The Overall, the prevalence of CNO has onset of CNO with higher acute phase proinflammatory state both in FMF pa- been estimated at between 1/160,000 reactants, lower haemoglobin concen- tients and asymptomatic MEFV carri- to 1/2,000,000 (27). Its pathogenesis trations, and a higher number of bone ers was related to the expression of oth- remains unclear; however, studies lesions. Moreover, they had a persis- er inflammatory diseases. Vasculitis, show that it is an autoinflammatory tent disease course more commonly JIA, spondyloarthritis, and IBD were disease like FMF. Mainly, IL-1β seems than patients without MEFV mutations the most commonly associated dis- to drive the disease (28). Additionally, that caused to add biologics to treat- eases among paediatric patients with IL-1, IL-6, and TNF-α were found in- ments of all patients with mutations. FMF (10, 23-25). In the current study, creased whereas a decreased IL-10 Our results demonstrating an increased two patients already had the diagno- secretion was identified (29). Autoanti- frequency of MEFV gene mutations in sis of FMF before the manifestations bodies generally not detected or found CNO and a more severe disease phe- of CNO. Three asymptomatic patients in low titers with a nonspecific profile notype of CNO in patients with MEFV were additionally found to have MEFV and the lack of antigen-specific T cells gene mutations support the possible mutations. Overall, this frequency endorse this hypothesis (29). Evidence relationship between CNO and MEFV was higher than the prevalence and supports that CNO might have a ge- gene mutations. We think that because the carrier rates of MEFV mutations netic basis especially after the identifi- of the infrequency of CNO and the in population-based field studies from cation of familial clustering cases with limited awareness about the disease, Turkey (0.1% and 14.8%, respectively CNO and CNO-associated diseases. this relationship might not have been (18, 19)). All patients with MEFV mu- Whole exome sequencing revealed realised until recently. Due to the pres- tations were found to have exon 10 FBLIM1 and FGR gene mutations in a ence of few paediatric rheumatologists mutations (four with M694V and one small portion of CNO patients as sus- in our country, probably, the patients with M680I) that were known to be ceptibility genes (4, 30). Additionally, with CNO had been treated with other associated with a severe phenotype of two monogenic diseases are defined as diagnoses. Likewise, we suggest that FMF (26). Moreover, the M694V allele syndromic CNOs. Majeed syndrome there should be more cases with CNO

S-116 Clinical and Experimental Rheumatology 2020 CNO with MEFV mutations / P.O. Avar.Aydin et al. in regions where FMF is prevalent as it 5. JANSSON A, RENNER ED, RAMSER J et al.: 21. TUNCA M, OZDOGAN H, KASAPCOPUR O et can be in our country. We believe that Classification of non-bacterial osteitis:- Ret al.: Familial Mediterranean Fever (FMF) in rospective study of clinical, immunological Turkey: Results of a nationwide multicenter particularly in these regions, the symp- and genetic aspects in 89 patients. Rheuma- study. Medicine (Baltimore) 2005; 84: 1-11. toms of FMF should be questioned tology 2007; 46: 154-60. 22. BRIK R, SHINAWI M, KASINETZ L, GERSHO- in detail in patients with CNO. In the 6. JANSSON AF, MÜLLER TH, GLIERA L et al.: NI-BARUCH R: The musculoskeletal mani- presence of FMF symptoms or early- Clinical score for nonbacterial osteitis in festations of familial Mediterranean fever in children and adults. Arthritis Rheum 2009; children genetically diagnosed with the dis- onset CNO with a severe disease phe- 60: 1152-9. ease. Arthritis Rheum 2001; 44: 1416-9. notype, MEFV gene mutations should 7. BORZUTZKY A, STERN S, REIFF A et al.: 23. AYDIN F, ÖZÇAKAR ZB, ÇAKAR N et al.: Sac- be performed. Pediatric chronic nonbacterial osteomyelitis. roiliitis in children with familial Mediterrane- Limitations of the current study are Pediatrics 2012; 130: e1190-7. an fever. J Clin Rheumatol 2019; 25: 69-73. 8. BEN-CHETRIT E, LEVY M: Familial Mediter- 24. ÖZÇAKAR ZB, YALÇINKAYA F, ÇAKAR N the limited number of patients with ranean fever. Lancet 1998; 351: 659-64. et al.: MEFV mutations modify the clinical CNO restricted to a group of Turkish 9. LACHMANN HJ, ŞENGÜL B, YAVUZŞEN TU presentation of Henoch-Schönlein purpura. J children from a single centre and their et al.: Clinical and subclinical inflammation Rheumatol 2008; 35: 2427-9. in patients with familial Mediterranean fever 25. YALÇINKAYA F, ÖZÇAKAR ZB, KASAP- relatively short follow-up. Multicen- and in heterozygous carriers of MEFV muta- ÇOPUR Ö et al.: Prevalence of the MEFV tric international studies are needed to tions. Rheumatology 2006; 45: 746-50. gene mutations in childhood polyarteritis no- confirm our findings. Moreover, the 10. ÖZÇAKAR ZB, ÇAKAR N, UNCU N, ÇELIKEL dosa. J Pediatr 2007; 151: 675-8. additional analysis of other candidate BA, YALÇINKAYA F: Familial Mediterranean 26. GANGEMI S, MANTI S, PROCOPIO V et al.: fever-associated diseases in children. QJM Lack of clear and univocal genotype-pheno- genes associated with CNO might have 2017; 110: 287-90. type correlation in familial Mediterranean fe- been beneficial. Despite these limita- 11. BEN-CHETRIT E, PELEG H, AAMAR S, HEY- ver patients: A systematic review. Clin Genet tions, the current study demonstrated MAN SN: The spectrum of MEFV clinical 2018; 94: 81-94. an increased frequency of MEFV gene presentations--is it familial Mediterranean 27. GIRSCHICK H: Chronic nonbacterial os- fever only? Rheumatology (Oxford) 2009; teomyelitis/Chronic recurrent multifocal mutations among patients with CNO 48: 1455-9. osteomyelitis [Internet]. 2013. Available and compared the differences between 12. BADER-MEUNIER B, VAN NIEUWENHOVE E, from: https://www.orpha.net/consor/cgi- CNO patients with and without MEFV BRETON S, WOUTERS C: Bone involvement bin/Disease_Search.php?lng=EN&data_ gene mutations for the first time. in monogenic autoinflammatory syndromes. id=21817&Disease_Disease_Search_ Rheumatology 2018; 57: 606-18. diseaseGroup=259680&Disease_Disease_ In conclusion, CNO appears as a dis- 13. SHIMIZU M, TONE Y, TOGA A et al.: Colchi- Search_diseaseType=OMIM&Disease(s)/ ease associated with MEFV gene muta- cine-responsive chronic recurrent multifocal group of diseases=Chronic-nonbacterial-os- tions. Moreover, the presence of MEFV osteomyelitis with MEFV mutations: a vari- teomyelitis-Chronic-recurrent-multifocal-os mutations seems to affect the disease ant of familial Mediterranean fever? Rheu- 28. LUKENS JR, GROSS JM, CALABRESE C et al.: matology 2010; 49: 2221-3. Critical role for -independent phenotype of CNO. Physicians practic- 14. DAIA A, KINI V, TAHA R, EL-SHANTI H, FATH- IL-1β production in osteomyelitis. Proc Natl ing in regions where FMF is prevalent ALLA B: Chronic recurrent multifocal osteo- Acad Sci USA 2014; 111: 1066-71. should be aware that patients with FMF myelitis in a patient with familial Mediter- 29. HOFMANN SR, KAPPLUSCH F, MÄBERT K, or asymptomatic MEFV carriers may ranean fever. Pediatr Rheumatol 2015; 13 HEDRICH CM: The molecular pathophysiol- (Suppl. 1): P104. ogy of chronic non-bacterial osteomyelitis present with CNO and these patients 15. LEE JJY, STIMEC J, CAMERON B, LAXER RM: (CNO)—a systematic review. Mol Cell Pedi- may have a severe disease phenotype Concurrent chronic recurrent multifocal os- atr 2017; 4: 7. of CNO. Lastly, we suggest that FMF teomyelitis and familial Mediterranean fever. 30. ABE K, COX A, TAKAMATSU N et al.: Gain- should be included in CNO-associated A case report. Clin Exp Rheumatol 2018; 36 of-function mutations in a member of the Src (Suppl. 115): S151. family kinases cause autoinflammatory bone conditions. 16. GEZGIN YILDIRIM D, BEDIR DEMIRAĞ T, disease in mice and . Proc Natl Acad AKDULUM İ, YILMAZ G, BAKKALOĞLU S: Sci USA 2019; 116: 11872-7. References Chronic non-bacterial osteomyelitis coexist- 31. FERGUSON PJ, CHEN S, TAYEH MK et al.: 1. ZHAO Y, FERGUSON PJ: Chronic nonbacteri- ent with familial Mediterranean fever. Clin Homozygous mutations in LPIN2 are re- al osteomyelitis and chronic recurrent multi- Exp Rheumatol 2018; 36 (Suppl. 115): S150. sponsible for the syndrome of chronic recur- focal osteomyelitis in children. Pediatr Clin 17. YALÇINKAYA F, ÖZEN S, ÖZÇAKAR ZB et rent multifocal osteomyelitis and congenital North Am 2018; 65: 783-800. al.: A new set of criteria for the diagnosis of dyserythropoietic anaemia (Majeed syn- 2. COX AJ, FERGUSON PJ: Update on the genet- familial Mediterranean fever in childhood. drome). J Med Genet 2005; 42: 551-7. ics of nonbacterial osteomyelitis in humans. Rheumatology 2009; 48: 395-8. 32. AKSENTIJEVICH I, MASTERS SL, FERGUSON Curr Opin Rheumatol 2018; 30: 521-5. 18. OZEN S, KARAASLAN Y, OZDEMIR O et al.: PJ et al.: An autoinflammatory disease with 3. HOFMANN SR, MORBACH H, SCHWARZ T, Prevalence of juvenile chronic arthritis and deficiency of the interleukin-1-receptor- an RÖSEN-WOLFF A, GIRSCHICK HJ, HEDRICH familial Mediterranean fever in Turkey: a tagonist. N Engl J Med 2009; 360: 2426-37. CM: Attenuated TLR4/MAPK signaling in field study.J Rheumatol 1998; 25: 2445-9. 33. SALEHZADEH F, ANARI H, SARKHANLOO S: from patients with CRMO results 19. SOYLEMEZOGLU O, KANDUR Y, GONEN S et Idiopathic CRMO and MEFV gene variant in impaired IL-10 expression. Clin Immunol al.: Familial Mediterranean fever gene muta- alleles: is there any relationship? Case Rep 2012; 145: 69-76. tion frequencies in a sample Turkish popula- Rheumatol 2019; 2019: 9847867. 4. COX AJ, DARBRO BW, LAXER RM et al.: tion. Clin Exp Rheumatol 2016; 34 (Suppl. 34. ÇIÇEK SÖ, ŞAHIN N, KARAMAN ZF et al.: Recessive coding and regulatory mutations 102): S97-100. The possible relationship between familial in FBLIM1 underlie the pathogenesis of 20. BEN-CHETRIT E, TOUITOU I: Familial Medi- Mediterranean fever and chronic nonbacte- chronic recurrent multifocal osteomyelitis terranean fever in the world. Arthritis Care rial osteomyelitis. J Clin Rheumatol 2020 (CRMO). PLoS One 2017; 12: 1-22. Res 2009; 61: 1447-53. Jun 12 [Online ahead of print].

Clinical and Experimental Rheumatology 2020 S-117