Publique Hôpitaux de Paris, Hôpital Necker-Enfants-Malades, INSERM U768, and Université René Descartes Paris 5, IFR 94, Paris, France;

Transcription

1 ARTHRITIS & RHEUMATISM Vol. 62, No. 1, January 2010, pp DOI /art , American College of Rheumatology Long-Term Efficacy of the Interleukin-1 Receptor Antagonist Anakinra in Ten Patients With Neonatal-Onset Multisystem Inflammatory Disease/Chronic Infantile Neurologic, Cutaneous, Articular Syndrome Bénédicte Neven, 1 Isabelle Marvillet, 2 Celine Terrada, 3 Alice Ferster, 4 Nathalie Boddaert, 5 Vincent Couloignier, 2 Graziella Pinto, 2 Anne Pagnier, 6 Christine Bodemer, 2 Bahram Bodaghi, 3 Marc Tardieu, 7 Anne Marie Prieur, 2 and Pierre Quartier 1 Objective. Cryopyrin-associated periodic syndromes (CAPS) are a group of rare autoinflammatory diseases. Neonatal-onset multisystem inflammatory disease (NOMID)/chronic infantile neurologic, cutaneous, articular syndrome (CINCA syndrome) is the most severe phenotype, with fever, rash, articular manifestations, and neurologic and neurosensory involvement. CAPS are caused by mutations in CIAS1, the gene encoding NLRP3, which plays a critical role in interleukin-1 (IL-1) processing. Anakinra, an IL-1 receptor antagonist, has been shown to be an effective treatment; however, data on long-term efficacy and safety have been sparse. This study was undertaken to assess the long-term efficacy and safety of anakinra treatment in patients with NOMID/CINCA syndrome. Methods. We retrospectively analyzed the medical records of NOMID/CINCA syndrome patients referred 1 Bénédicte Neven, MD, Pierre Quartier, MD: Assistance Publique Hôpitaux de Paris, Hôpital Necker-Enfants-Malades, INSERM U768, and Université René Descartes Paris 5, IFR 94, Paris, France; 2 Isabelle Marvillet, MD, Vincent Couloignier, MD, PhD, Graziella Pinto, MD, Christine Bodemer, MD, PhD, Anne Marie Prieur, MD: Assistance Publique Hôpitaux de Paris, Hôpital Necker- Enfants-Malades, Paris, France; 3 Celine Terrada, MD, Bahram Bodaghi, MD, PhD: Assistance Publique Hôpitaux de Paris, Hôpitaux de la Pitié-Salpétrière, Paris, France; 4 Alice Ferster, MD, PhD: Hôpital des Enfants Reine Fabiola, Brussels, Belgium; 5 Nathalie Boddaert, MD, PhD: Assistance Publique Hôpitaux de Paris, Hôpital Necker- Enfants-Malades, Paris, France, and INSERM U797, Orsay, France; 6 Anne Pagnier, MD: CHU de Grenoble, Grenoble, France; 7 Marc Tardieu, MD: Assistance Publique Hôpitaux de Paris, CHU de Bicètre, Paris, France. Address correspondence and reprint requests to Bénédicte Neven, MD, Hôpital Necker, 149 Rue de Sevres, Paris 75015, France. benedicte.neven@nck.aphp.fr. Submitted for publication May 11, 2009; accepted in revised form September 25, to 2 centers, who had started anakinra treatment before June Results. There were 10 patients with NOMID/ CINCA syndrome who had been treated with anakinra. The patients ages at the time anakinra treatment was initiated ranged from 3 months to 20 years. They had been followed up for months. Sustained efficacy in the treatment of systemic inflammation and, in some cases, neurologic involvement and growth parameters, was achieved. The dosage of anakinra required for efficacy ranged from 1 to 3 mg/kg/day in the 8 oldest patients and from 6 to 10 mg/kg/day in the 2 youngest. Residual central nervous system inflammation and deafness persisted in some patients, especially if there had been a delay in diagnosis and treatment. Secondary amyloidosis persisted in cases in which it was present at treatment initiation, but no new lesions developed. No effect on overgrowth arthropathy was observed. Adverse events consisted of mild injection-site reactions. Conclusion. The present results indicate that anakinra treatment is effective over the long term in NOMID/CINCA syndrome. However, treatment has to be initiated before irreversible lesions develop, and, particularly in very young patients, dosage adjustment is required. Cryopyrin-associated periodic syndromes (CAPS) are a group of rare autoinflammatory diseases including familial cold autoinflammatory syndrome (FCAS) (MIM ), Muckle-Wells syndrome (MWS) (MIM ), and chronic infantile neurologic, cutaneous, articular syndrome (CINCA syndrome) (MIM ) (also known as neonatal-onset multisystem inflammatory disease [NOMID]). These syndromes, 258

2 ANAKINRA EFFICACY IN NOMID/CINCA SYNDROME 259 initially described as different entities, have overlapping symptoms, i.e., fever, urticaria-like skin rash, and articular involvement, associated with increases in the levels of biologic markers of polymorphonuclear neutrophil (PMN) driven inflammation. They represent a continuum of severity, with FCAS being the mildest condition, MWS being of intermediate severity, and NOMID/ CINCA syndrome being the most severe disease, with patients exhibiting a broader range of symptoms. In NOMID/CINCA syndrome, the course of the disease is more chronic than recurrent. The initial symptom is generally an urticaria-like rash, occurring very early in life. Fever may be mild. The severity of bone and joint involvement is variable: approximately two-thirds of patients develop arthralgia and transient swelling during flares, whereas the remaining one-third develop severe and disabling arthropathy caused by overgrowth of the patella and epiphyses of long bones, resulting in gross deformity of the joints. Central nervous system (CNS) abnormalities are observed in almost all patients, due to chronic aseptic meningitis, with PMNs infiltrating the cerebrospinal fluid (CSF). Chronic headaches and papilledema are frequently observed, due to chronically increased intracranial pressure. Severely affected patients may display cognitive impairment. Neurosensory involvement is also observed. Ocular disease, i.e., anterior or posterior uveitis or papillitis, may occur. Optic atrophy may develop, with progression to blindness. Perceptive deafness is frequent, with onset during childhood or adulthood. AA amyloidosis develops in 25% of patients. Severe disability and premature death are possible. Leukocytosis and increases in levels of acute-phase reactants are permanent in most cases (1). CAPS are all caused by dominant inherited or de novo mutations in CIAS1, the gene encoding NLRP3 (also known as cryopyrin/nalp3/pypaf1) (2 4). NLRP3 is a component of the interleukin-1 (IL-1) inflammasome regulating the production of IL-1, a proinflammatory cytokine. Results of in vitro studies have suggested that CIAS1 mutations may cause CIAS1- related diseases through a gain-of-function effect (5,6). Advances in the understanding of the genetic basis and mechanisms of these disorders have led to the development of new treatments targeting the IL-1 pathway, such as anakinra, the recombinant nonglycosylated homolog of human IL-1 receptor antagonist, which competitively inhibits the binding of IL-1 and IL-1 to the IL-1 receptor. Previous studies have demonstrated that anakinra is effective in the treatment of CIAS1-associated diseases. It rapidly controls constitutional symptoms, urticaria-like rash, and acute-phase reactant levels in all of these diseases (7 15). Some studies have demonstrated improvements in hearing and amyloid-related proteinuria in selected cases of MWS (16). Golbach- Mansky et al also demonstrated the efficacy of anakinra in the treatment of CNS symptoms and neurosensory involvement in a series of 18 patients with NOMID/ CINCA syndrome, after 6 months of therapy (15). To date, however, there have been no published reports of studies evaluating the long-term effects of anakinra treatment in these patients. PATIENTS AND METHODS We retrospectively analyzed the efficacy and safety of anakinra treatment in NOMID/CINCA syndrome patients whose treatment was initiated before June 2007 at 2 pediatric centers (Hôpitaux Necker Enfants-Malades and Hôpital des Enfants Reine Fabiola). Patients and parents were informed that anakinra treatment had been approved only for adult patients with rheumatoid arthritis. Oral informed consent was obtained before treatment was initiated. Medical history was recorded; the frequency and severity of clinical symptoms such as rash, fever, arthralgia or arthritis, and headache before treatment and during followup were noted. Inflammatory biologic markers, such as acute-phase reactants (serum amyloid A [SAA], C-reactive protein [CRP], and erythrocyte sedimentation rate [ESR]) and PMN count were measured regularly. Results of the following tests, if performed, were also noted: audiography, vision evaluations (funduscopy, angiography, visual field, anterior chamber evaluation by laser flare photometry), CSF examination (protein level, cell counts), and tests of cognitive function (Wechsler Preschool and Primary Scale of Intelligence, or Wechsler Intelligence Scale for Children). Brain magnetic resonance imaging (MRI) (1.5Tesla; T1-and T2-weighted fluid-attenuated inversion recovery [FLAIR] imaging and contrast injection) was performed before and after treatment in most patients, and the results were retrospectively reviewed by the same experienced radiologist. The end point of the study for this analysis was January Fisher s exact test was used to compare categorical variables. P values (2-tailed) less than or equal to 0.05 were considered significant. Statistical analyses were conducted with Prism 5 software (GraphPad Software, San Diego, CA). RESULTS Characteristics of the patients before anakinra treatment. Ten patients (3 male, 7 female) were studied. Their followup (from the initiation of anakinra treatment to the end point of the study in January 2009) ranged from 26 months to 42 months. All patients fulfilled clinical criteria for NOMID/CINCA syndrome (urticaria-like skin rash, CNS involvement, and/or overgrowth arthropathy) (1). The diagnosis was confirmed genetically in all patients except one (patient 6). The

3 260 NEVEN ET AL Table 1. Characteristics of the patients with neonatal-onset multisystem inflammatory disease/chronic infantile neurologic, cutaneous, articular syndrome at the time of initiation of anakinra therapy Patient Age, years Rash* Fever* Headache* Overgrowth arthropathy Mental delay Papilledema Aseptic meningitis Hearing loss Low BMI CIAS1 mutation M662T T436N D303N 4 16 I334V T436N V351M 8 6 H358R T436P D303H * occasional; weekly; daily. absent; present (aseptic meningitis was defined as 5 cells/mm 3 and/or cerebrospinal fluid protein level 20 gm/liter; low body mass index [BMI] was defined as a BMI below 2.0). mild (for hearing loss, db); moderate (for hearing loss, db); severe (for hearing loss, 70 db). Ventriculoperitoneal shunt inserted at age 2 years, due to major intracranial hypertension and macrocrania. characteristics of the patients before treatment are shown in Table 1. Patients 9 and 10 were very young (3 months and 4 months, respectively, at treatment initiation), and they are considered separately from the other 8 patients. The median age of patients 1 8 was years (range ) when anakinra treatment was started. A positive family history was recorded for 3 patients (including 2 siblings [patients 2 and 5]). All patients exhibited symptoms from the first days of life; 5 were born preterm and/or small for gestational age. All patients had previously received several treatments, e.g., nonsteroidal antiinflammatory drugs, etanercept, cyclophosphamide, chlorambucil, thalidomide, and colchicine. Patients 1, 4, and 6 were receiving steroid treatment (prednisolone mg/kg/day) when anakinra therapy was initiated. All patients had active disease, with frequent episodes of fever, urticaria-like skin rash, and frequent episodes of pain. All had experienced daily to weekly episodes of arthralgia, and 5 had experienced acute episodes of arthritis in the past, although none had acute arthritis at the time of anakinra initiation. Only patient 6 exhibited overgrowth arthropathy. Clinical CNS involvement and aseptic meningitis were present in all patients (Table 1). Five patients exhibited developmental retardation, with IQ scores between 50 and 70. Brain MRI (with FLAIR imaging and contrast injection) performed before treatment in all patients except patient 7 showed abnormalities, consisting of abnormal enhancement of small vessels of the basal ganglia after contrast injection (n 7) and periventricular white matter lesions seen on FLAIR sequences (n 4) (Figures 1 and 2). Ventriculomegaly was also noted in 3 patients (associated with cerebral atrophy in 2). Seven patients had abnormal audiography results, with percep- Figure 1. Axial magnetic resonance imaging in 2 patients with neonatal-onset multisystem inflammatory disease/chronic infantile neurologic, cutaneous, articular syndrome before and after initiation of anakinra. A, T1 spin-echo weighted image after contrast injection in patient 8 before anakinra treatment, showing small vessel enhancement at the level of basal ganglia (striatum) (arrow). B, T1 spin-echo weighted images before (left) and after (right) contrast injection in patient 8 after 3 years of anakinra treatment with a progressive increase in dosage up to 3 mg/kg/day, showing no small vessel enhancement at the level of the striatum. C, T1 spin-echo weighted image after contrast injection in patient 1 before anakinra treatment, showing small vessel enhancement at the level of basal ganglia (striatum) (arrow). D, T1 spin-echo weighted image after contrast injection in patient 1 after 3 years of anakinra treatment with a progressive increase in dosage up to 2 mg/kg/day, showing attenuation of small vessel enhancement at the level of the striatum (indicative of improvement).

4 ANAKINRA EFFICACY IN NOMID/CINCA SYNDROME 261 Figure 2. Coronal fluid-attenuated inversion recovery (FLAIR) magnetic resonance imaging (MRI) in 3 patients with neonatal-onset multisystem inflammatory disease/chronic infantile neurologic, cutaneous, articular syndrome. A, FLAIR MRI before anakinra treatment in patients 4 (left), 2 (middle), and 1 (right), showing periventricular white matter abnormalities (arrows). B and C, Improvement in the lesions (arrows) in patient 4 after (C) compared with before (B) 2 years of treatment with anakinra (initially at 1 mg/kg/day, with the dosage increased after 1 year to 2 mg/kg/day because of residual headaches and papilledema, both of which resolved after dosage adjustment). tive deafness, and had been experiencing hearing loss for 2 10 years. Of patients 1 8, 7 underwent extensive ophthalmologic evaluation (all at the same center). Patients 2 and 8 had visual impairment due to partial optic atrophy, but none of the patients exhibited active inflammation of the anterior and posterior chambers. Patient 7 had recurrent episcleritis. Visual field studies were performed in 4 patients, all of whom exhibited abnormalities, with mild retinal blind spot enlargement. In all patients, height was below 2 SD (between 7.5 and 2.3 SD; median 3.5 SD) and in 2, body mass index (BMI) was below 2 SD. Patients 1, 2, and 4 had completed puberty at baseline; patient 2 had secondary amenorrhea. Puberty was delayed in patient 3, who displayed no signs of puberty at the age of 16.8 years. Amyloidosis was present in patients 3 and 5. Patient 3 had mildly impaired renal function (glomerular filtration rate 65 ml/minute) but no proteinuria, at the time of treatment initiation. Patient 5 presented with significant proteinuria (8 gm/24 hours), with nephrotic syndrome (albuminemia 11 gm/liter) but normal renal function. Renal biopsies in these 2 individuals revealed intense AA amyloid deposition in both, and major glomerular sclerosis in patient 3. All patients had high levels of acute-phase reactants. The median CRP concentration, ESR, and SAA concentration before the start of treatment was 69 mg/liter (range ), 62 mm/hour (range ), and 150 mg/liter (48 386), respectively. Neutrophilia was also observed in all patients (median count 12,200/ l [range 6,700 23,700]). Patients 9 and 10 were 3 months old and 4 months old, respectively, when anakinra treatment was started. Both presented with severe disease. Patient 9 was born at 31.6 weeks gestational age, with normal weight and height for gestational age. She presented, at birth, with rash, fever, aseptic meningitis, and high levels of inflammation markers (CRP 150 mg/liter). Patient 10 was born at 32.5 weeks of gestational age and had low birth weight and height for gestational age (1,550 gm and 37 cm; both below the third percentile). She presented, at birth, with aseptic peritonitis, urticarial rash, fever, aseptic meningitis, and high levels of inflammation markers (CRP 305 mg/liter). Both patients exhibited active disease at the time anakinra treatment was initiated. Both had low weight ( 3 SD and 3.7 SD in patients 9 and 10, respectively) and height ( 2 SDand 4.3 SD, respectively). Levels of inflammation markers (CRP, ESR, and neutrophils) were very high. CSF parameters were abnormal (pleiocytosis 140/ l and 26/ l, respectively, and CSF protein levels 156 mg/liter and 80 mg/liter, respectively). Findings of ophthalmologic evaluation (funduscopy and visual evoked potential) and hearing evaluation (auditory evoked potential) were normal. Brain MRI with contrast injection was performed before treatment in patient 9, and showed abnormal enhancement of small vessels of the basal ganglia; FLAIR sequences were not performed. Response to anakinra. In all patients, anakinra treatment was initiated at a low dosage (1 mg/kg/day). The daily dosage was gradually increased in 6 of the 8 older patients (patients 1 8), to 2 mg/kg (n 4) or 3 mg/kg (n 2) after a minimum of 6 months of treatment. The main reasons for dosage adjustment were persistent neurologic involvement (headaches and/or papilledema and/or abnormal CSF examination) (n 5) and persistent intermittently high levels of biologic inflammation markers (n 3). In patients 9 and 10, daily dosages were increased to 6 mg/kg and 10 mg/kg, respectively, because of persistent clinical symptoms and high levels of biologic inflammation markers. Clinical response to treatment. In patients 1 8, the median time of followup since the initiation of treatment was 31 months (range months). All patients displayed a rapid clinical response to anakinra, with complete remission of rash, fever, arthralgia, and myal-

5 262 NEVEN ET AL Figure 3. Changes in levels of biologic markers of inflammation (C-reactive protein [CRP] and serum amyloid A [SAA]) and cerobrospinal fluid (CSF) parameters with anakinra treatment in patients with neonatal-onset multisystem inflammatory disease/chronic infantile neurologic, cutaneous, articular syndrome. A, CRP concentration in patients 1 10 by duration of treatment (maximum daily dosage of anakinra at last followup was 1 mg/kg in patients 2 and 3, 2 mg/kg in patients 1, 4, 5, and 6, 3 mg/kg in patients 7 and 8, 6 mg/kg in patient 9, and 10 mg/kg in patient 10). B and C, CRP and SAA concentrations in patients 1 8 by daily dosage of anakinra. D, CRP concentration in patients 9 and 10 by daily dosage of anakinra. E and F, CSF cellularity and protein levels in all patients (except patient 5, who refused further lumbar puncture after baseline, and patient 7, who had a ventriculoperitoneal shunt) at baseline and last evaluation. The shaded area in F represents the normal range of CSF protein levels. gia within 24 hours of treatment initiation at 1 mg/kg/ day. This improvement persisted over time, even with low doses of anakinra, except in patient 8, who had a disease flare with fever and rash 1 year after treatment initiation; clinical remission was achieved by increasing the daily dosage of anakinra to 3 mg/kg. The overgrowth arthropathy observed in patient 6 did not improve with treatment. A comparison of knee radiographs at baseline (treatment initiation) and at last followup showed a slow progression of abnormalities and size of metaphyses and epiphyses over a 3-year period. The 2 youngest patients (patients 9 and 10) were treated for 34 months and 30 months, respectively. Symptoms improved with anakinra at 1 mg/kg/day, but the rash rapidly reappeared 6 8 hours after injections. The dosage was rapidly increased to 2 mg/kg/day in 2 daily injections, and eventually to 6 mg/kg/day in patient 9 and 10 mg/kg/day in patient 10, to control disease activity. Even with these higher doses, patient 10 experienced episodic rashes during viral infections. Biologic markers of inflammation during treatment. In patients 1 8, the levels of all inflammation markers (CRP, SAA, ESR, and neutrophils) rapidly decreased 1 month after the initiation of anakinra (Figure 3A). Further followup revealed dose-dependent benefits in 3 patients, as illustrated in Figures 3B and C for CRP and SAA. The PMN count and ESR followed a similar pattern (data not shown). Inflammation marker levels also significantly and persistently improved in patients 9 and 10 after adjustment of the anakinra dosage (Figure 3D).

6 ANAKINRA EFFICACY IN NOMID/CINCA SYNDROME 263 Table 2. Changes in neurologic involvement with anakinra treatment* Before anakinra Last followup Cerebral MRI anomalies Cerebral MRI anomalies Patient Aseptic meningitis Papilledema White matter change Small vessel basal ganglia VM/CA Daily dosage, mg/kg Aseptic meningitis Papilledema White matter change Small vessel basal ganglia VM/CA 1 / 2 Improved Improved Improved / 2 / 1 Stable NA NA NA 3 / 1 Improved / 4 / 2 Improved Improved Improved Stable 5 / 2 NA NA NA NA 6 / 2 Stable NA NA NA 7 NA NA NA 3 NA NA NA 8 / 3 Improved Stable 9 NA / 6 NA NA NA 10 NA NA / 10 NA NA NA * Cerebral magnetic resonance imaging (MRI) anomalies included periventricular white matter abnormality seen on fluid-attenuated inversion recovery sequence and abnormal enhancement of small vessels in basal ganglia after gadolinium injection. VM/CA ventriculomegaly/cerebral atrophy; NA not assessed. absent; present. mild; moderate; severe. Ventriculoperitoneal shunt inserted at age 2 years, due to major intracranial pressure and macrocrania. Neurologic involvement. Patients 1 8 all exhibited CNS involvement at baseline (headaches [n 7] and/or papilledema [n 7] and/or abnormalities found on CSF examination [n 8]). After 6 months of treatment with anakinra at 1 mg/kg/day, headaches had completely resolved in 3 patients and significantly decreased in frequency and severity in the remaining 4. CSF was reanalyzed in 6 patients. White blood cell counts had significantly decreased in all of these patients compared with baseline (P 0.045). Protein levels had decreased, but not significantly (data not shown). One patient (patient 3) exhibited complete normalization of all CSF parameters. Papilledema normalized in 3 patients and improved in the other 4. Persistence of headaches was associated with persistence of papilledema, suggesting a correlation between these symptoms. In 5 patients with persistent headaches and/or persistent papilledema after 1 year of treatment (patients 1, 4, 6, 7, and 8), the dosage of anakinra was gradually increased from 1 mg/kg/day to 3 mg/kg/day in order to control CNS involvement. Symptoms resolved, apart from episodic headaches in 1 patient; however, CSF parameters normalized in only 1 patient, improved in 2 patients, and remained stable in 2. A reevaluation of mental retardation at the last followup showed no improvement. Cerebral MRI was reassessed 2 3 years after treatment initiation, after optimization of the anakinra dosage, in 4 patients. The results are presented in Figures 1 and 2 and Table 2. Small vessel enhancement after contrast injection, observed in the 4 patients at baseline, improved, as did periventricular white matter anomalies initially present in 2 of them. The irritability displayed by patients 9 and 10 at baseline rapidly improved with treatment. No papilledema developed during followup, and both patients exhibited a rapid improvement in CSF parameters within 3 months and complete normalization after 12 months. Patient 9 developed mild developmental retardation. CSF cellularity and protein content were compared in all patients who had analyzable assessments at baseline and at the last followup, after the anakinra dosage had been adjusted to control headaches and papilledema (i.e., all patients except patients 5 and 7). Significant improvement in these parameters after treatment was observed (Figures 3E and F). Neurosensory involvement. Seven of the 8 oldest patients reported hearing loss at baseline. Patients 1 and 3 exhibited a 15-dB improvement on audiography performed 6 months after the start of anakinra treatment. This improvement remained stable over time. In 4 patients, hearing loss remained stable even after the dosage of anakinra had been increased. Visual acuity remained stable in all patients. The visual field defects present in the 4 patients tested at baseline normalized in all 4, together with resolution of papilledema. No uveitis or papillitis developed. Neither patient 9 nor patient 10

7 264 NEVEN ET AL Figure 4. Growth parameters before and during anakinra treatment in patients with neonatal-onset multisystem inflammatory disease/chronic infantile neurologic, cutaneous, articular syndrome. A, Changes in height over time in 7 patients with growth potential. Patient 6 had overgrowth arthropathy and was small for age before treatment and 3 years after the initiation of treatment; however, his growth velocity normalized. B, Changes in body mass index (BMI) over time in all patients. not all patients reached 36 months of followup. Among patients who reached 36 months of followup, the median height and BMI, respectively, were 3.36 SD and 1.54 SD. exhibited neurosensory involvement at baseline or during followup. Growth parameters. Growth could be assessed in only 5 of the 8 older patients (patients 3, 5, 6, 7, and 8 [in whom anakinra treatment was begun before completion of pubertal development]). Median growth increased from 3 cm/year (range ) before treatment to 7 cm/year (range 4 7.5) after 1 year of anakinra treatment and 7 cm/year (range 2 8.5) after 2 years (Figure 4A). All patients gained height and exhibited increased BMI (Figures 4A and B). Two patients eventually became overweight, with BMI more than 2 SD. Patient 3 exhibited delayed pubertal development at baseline, but puberty rapidly developed within 6 months of treatment. Secondary amenorrhea also resolved within 3 months in patient 1. Patient 9 had short stature and low BMI, which persisted through the last followup. Patient 10 s height improved, and her BMI normalized (Figures 4A and B). Amyloidosis. Two patients had amyloidosis at baseline. One (patient 3) had an abnormal filtration rate, which did not improve after 38 months of anakinra treatment, although the SAA concentration had normalized within 6 months. However, no proteinuria developed. In the other (patient 5), proteinuria, which was present at baseline, gradually improved from 8.5 gm/24 hours to 1.5 gm/24 hours at 6 months and 0.45 gm/24 hours at 1 year of anakinra treatment. The SAA level became normal after 6 months of treatment and albuminemia normalized after 1 year. Safety and tolerability. No adverse effects were observed, except mild local stinging and erythema at the injection sites, which gradually decreased in most patients. No severe infection occurred. All patients had been optimally immunized against Streptococcus pneumoniae and Haemophilus influenzae. Patients 9 and 10 received daily antibiotic prophylaxis with penicillin and trimethoprim, due to the high doses of anakinra administered and their very young age. DISCUSSION We studied 10 patients with NOMID/CINCA syndrome, the most severe form of CAPS, who were treated with anakinra for months. Our findings confirmed remarkable long-term efficacy of anakinra in treating a broad range of symptoms (including systemic manifestations, neurologic involvement, and growth parameters). The magnitude of the response was dose dependent. Older patients ( 6 years) (patients 1 8) and younger patients ( 2 years) (patients 9 and 10) were considered separately. In the 2 youngest patients, particularly high doses of anakinra were needed to control disease activity. Usual dosages of anakinra (1 mg/kg/day) were sufficient to control constitutional symptoms, such as

8 ANAKINRA EFFICACY IN NOMID/CINCA SYNDROME 265 fever, rash, and arthralgia, in 7 of the 8 patients who were over the age of 6 years at baseline. However, in 6 of them, higher dosages (up to 2 or 3 mg/kg/day) were needed in order to achieve persistent normalization of biologic inflammation marker levels and decreases in CNS inflammation. These dosages were higher than those usually recommended for FCAS and MWS (1 mg/kg/day, or mg/kg/day if the disease is mild) (14) and the dosages reported in previous studies of patients with NOMID/CINCA syndrome (15,17). After optimization of the anakinra dosage, headaches and papilledema resolved simultaneously in all patients. However, residual CNS inflammation persisted in some patients, as evidenced by residual hyperleukocytosis and/or high protein levels on CSF examination. Residual periventricular white matter abnormalities on cerebral MRI (suggesting abnormal transependymal resorption of CSF) and abnormal enhancement of small vessels of the basal ganglia (suggesting inflammation of vessel walls) also indicated residual activity of the disease. Even higher dosages of anakinra might be needed in order to abolish CNS inflammation completely, but the benefits should be balanced against potential risks. The high dosages required in patients with NOMID/ CINCA syndrome were presumably due to the high degree of severity of the NOMID/CINCA phenotype among CIAS1-related diseases. The drug may also have lower bioavailability in the CNS due to the blood brain barrier. Pharmacokinetics studies of anakinra in the CSF would be of interest to confirm this hypothesis. Among the group of patients over the age of 6 years at baseline, those with growth potential gained considerably in height and displayed a large increase in growth velocity. They also exhibited a marked improvement in BMI. The rapid maturation of secondary sexual traits was observed in 1 patient with delayed pubertal development, and rapid resolution of secondary amenorrhea was observed in another. These improvements in growth and development parameters presumably reflect the resolution of a previous long-lasting catabolic state related to chronic IL1- driven inflammation. In the present series, neurosensory abnormalities persisted with anakinra treatment in most of the patients who were older than 6 years at baseline. We observed a mild improvement of hearing in only 2 of 7 patients, with stabilization of hearing loss in the others. The etiology of hearing loss in CIAS1-related diseases remains unclear. The improvement of hearing with anti IL-1 treatment in some patients and the abnormal cochlear enhancement observed on MRI with gadolinium injection and in FLAIR sequences described in some patients (15) strongly support the notion of an inflammatory origin. However, necropsy studies in MWS patients with deafness showed anatomic abnormalities of the inner ear, which lacked a Corti organ and displayed atrophy of the cochlear nerve. This finding suggests that persistent inflammation of the cochlea may damage neuroepithelial Corti hair cells and the cochlear nerve, making hearing loss irreversible. Anakinra treatment may have been initiated too late in our patients, after irreversible damage had already occurred. Similarly, the impairment of visual acuity due to optic atrophy observed in 2 patients stabilized, but did not improve, with treatment. This impairment is an irreversible end-stage consequence of chronic excessive intracranial pressure and chronic papillitis. Mental retardation, resulting from the destruction of cerebral tissue by chronic inflammation, also remained stable under treatment. These findings highlight the need to initiate treatment as soon as possible, before the onset of irreversible damage caused by chronic inflammation. Only 1 patient in our series had developed progressive overgrowth arthropathy of the knees, beginning in the first year of life. A comparison of knee radiographs obtained at baseline and at last followup showed a slow progression of the size and irregularities of metaphyses and epiphyses. This observation, in accordance with other reports, suggests that overgrowth arthropathy is not driven by IL-1. As NALP3 is expressed in cartilage (3,18), such overgrowth may be due to abnormal apoptosis at the site of enchondral ossification. Patients 9 and 10, ages 3 4 months at baseline, manifested severe disease (preterm birth, neonatal systemic manifestations, aseptic meningitis, and high levels of biologic markers of inflammation). Treatment was remarkably beneficial in these patients; however, a higher dosage was required (6 mg/kg/day and 10 mg/kg/ day, respectively, administered in 2 injections per day). Systemic manifestations of the disease and biologic markers of inflammation were persistently controlled. Irritability and CSF parameters normalized, suggesting good control of neurologic involvement. Both of these patients remained free of neurosensory involvement. Nevertheless, prolonged followup is required in order to confirm the ability of anakinra to confer long-term protection against deafness and visual impairment. Neither patient developed overgrowth arthropathy. Benefits in terms of growth were limited in patient 9, who remained short for her age and retained a BMI of 2 SD below the normal mean. Patient 10 remained short for her age despite the normalization of BMI, presumably because of her small height and weight for gestational

9 266 NEVEN ET AL age. The unusually high dosage of anakinra required in these 2 patients may be accounted for by the severity of the disease and, probably, also by the more rapid clearance of the drug in young children. Several IL-1 inhibitors are currently under development. These molecules include cytokine traps and monoclonal antibodies against IL-1. The safety and efficacy of weekly subcutaneous injections (160 mg/ week) of rilonacept have recently been evaluated in patients with FCAS and MWS (19,20). The efficacy of treatment was confirmed in a 1-year study (19). Rilonacept has recently been approved by the US Food and Drug Administration for the treatment of MWS and FCAS in patients over the age of 11 years. Canakinumab, a new human IgG1 monoclonal antibody that is directed against IL-1 and has a long half-life, is currently being investigated for use in CAPS. However, the efficacy of these new drugs in NOMID/CINCA syndrome patients with neurologic involvement is unknown. IL-1 blockade, particularly by long-acting molecules with high affinity, entails a potential risk of serious infections, particularly in young children with immature immune responses against encapsulated bacteria, and safety data on these long-acting molecules are currently sparse. Optimizing vaccination against S pneumoniae and use of antibiotic prophylaxis in selected cases should be considered until better data on the safety of anakinra in young children become available. In conclusion, CIAS1-related diseases provide an example of the effective translation of basic science to patient care. Studies of the mechanisms underlying these diseases demonstrated a fundamental role of IL-1. This series of patients with NOMID/CINCA syndrome treated with anakinra, a homolog of human IL-1 receptor antagonist, highlights the benefits and limitations of this drug in the long term. The treatment of NOMID/ CINCA syndrome is more complex than that of milder forms of CAPS, due to the broader range of symptoms. It may be difficult to achieve complete resolution of neurologic involvement, and high doses are often required. Treatment should be initiated as soon as possible, before damage due to chronic inflammation becomes irreversible. The prenatal onset of the disease observed in some patients is a matter of potential concern. Monitoring and prompt adjustment of anakinra dosages in infants and young children is recommended. ACKNOWLEDGMENTS We are grateful to Ms Isabelle Funck-Brentano for performing neuro-developmental assessments, and to Dr. Genevieve de Saint Basile (INSERM U768) and Ms Laurence Cuisset (CHU Cochin, Paris, France) for performing CIAS1 genotyping. AUTHOR CONTRIBUTIONS All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Neven had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study conception and design. Neven, Boddaert, Prieur, Quartier. Acquisition of data. Neven, Marvillet, Ferster, Boddaert, Couloignier, Pinto, Pagnier, Bodemer, Bodaghi, Tardieu, Prieur, Quartier. Analysis and interpretation of data. Neven, Marvillet, Terrada, Ferster, Boddaert, Couloignier, Pinto, Bodaghi, Tardieu, Prieur, Quartier. REFERENCES 1. Prieur AM. A recently recognised chronic inflammatory disease of early onset characterised by the triad of rash, central nervous system involvement and arthropathy. Clin Exp Rheumatol 2001; 19: Hoffman HM, Mueller JL, Broide DH, Wanderer AA, Kolodner RD. Mutation of a new gene encoding a putative pyrin-like protein causes familial cold autoinflammatory syndrome and Muckle- Wells syndrome. Nat Genet 2001;29: Feldmann J, Prieur AM, Quartier P, Berquin P, Certain S, Cortis E, et al. Chronic infantile neurological cutaneous and articular syndrome is caused by mutations in CIAS1, a gene highly expressed in polymorphonuclear cells and chondrocytes. Am J Hum Genet 2002;71: Aksentijevich I, Nowak M, Mallah M, Chae JJ, Watford WT, Hofmann SR, et al. De novo CIAS1 mutations, cytokine activation, and evidence for genetic heterogeneity in patients with neonatal-onset multisystem inflammatory disease (NOMID): a new member of the expanding family of pyrin-associated autoinflammatory diseases. Arthritis Rheum 2002;46: Dowds TA, Masumoto J, Zhu L, Inohara N, Nunez G. Cryopyrininduced interleukin 1 secretion in monocytic cells: enhanced activity of disease-associated mutants and requirement for ASC. J Biol Chem 2004;279: Agostini L, Martinon F, Burns K, McDermott MF, Hawkins PN, Tschopp J. NALP3 forms an IL-1 -processing inflammasome with increased activity in Muckle-Wells autoinflammatory disorder. Immunity 2004;20: Hawkins PN, Lachmann HJ, McDermott MF. Interleukin-1-receptor antagonist in the Muckle-Wells syndrome. N Engl J Med 2003;348: Hawkins PN, Lachmann HJ, Aganna E, McDermott MF. Spectrum of clinical features in Muckle-Wells syndrome and response to anakinra. Arthritis Rheum 2004;50: Hoffman HM, Rosengren S, Boyle DL, Cho JY, Nayar J, Mueller JL, et al. Prevention of cold-associated acute inflammation in familial cold autoinflammatory syndrome by interleukin-1 receptor antagonist. Lancet 2004;364: Frenkel J, Wulffraat NM, Kuis W. Anakinra in mutation-negative NOMID/CINCA syndrome: comment on the articles by Hawkins et al and Hoffman and Patel. Arthritis Rheum 2004;50: Granel B, Serratrice J, Disdier P, Weiller PJ. Dramatic improvement with anakinra in a case of chronic infantile neurological cutaneous and articular (CINCA) syndrome. Rheumatology (Oxford) 2005;44: Kallinich T, Hoffman HM, Roth J, Keitzer R. The clinical course of a child with CINCA/NOMID syndrome improved during and

10 ANAKINRA EFFICACY IN NOMID/CINCA SYNDROME 267 after treatment with thalidomide. Scand J Rheumatol 2005;34: Lovell DJ, Bowyer SL, Solinger AM. Interleukin-1 blockade by anakinra improves clinical symptoms in patients with neonatalonset multisystem inflammatory disease. Arthritis Rheum 2005;52: Leslie KS, Lachmann HJ, Bruning E, McGrath JA, Bybee A, Gallimore JR, et al. Phenotype, genotype, and sustained response to anakinra in 22 patients with autoinflammatory disease associated with CIAS-1/NALP3 mutations. Arch Dermatol 2006;142: Goldbach-Mansky R, Dailey NJ, Canna SW, Gelabert A, Jones J, Rubin BI, et al. Neonatal-onset multisystem inflammatory disease responsive to interleukin-1 inhibition. N Engl J Med 2006;355: Alexander T, Klotz O, Feist E, Ruther K, Burmester GR, Pleyer U. Successful treatment of acute visual loss in Muckle-Wells syndrome with interleukin 1 receptor antagonist. Ann Rheum Dis 2005;64: Matsubara T, Hasegawa M, Shiraishi M, Hoffman HM, Ichiyama T, Tanaka T, et al. A severe case of chronic infantile neurologic, cutaneous, articular syndrome treated with biologic agents. Arthritis Rheum 2006;54: McCall SH, Sahraei M, Young AB, Worley CS, Duncan JA, Ting JP, et al. Osteoblasts express NLRP3, a nucleotide-binding domain and leucine-rich repeat region containing receptor implicated in bacterially induced cell death. J Bone Miner Res 2008;23: Goldbach-Mansky R, Shroff SD, Wilson M, Snyder C, Plehn S, Barham B, et al. A pilot study to evaluate the safety and efficacy of the long-acting interleukin-1 inhibitor rilonacept (interleukin-1 Trap) in patients with familial cold autoinflammatory syndrome. Arthritis Rheum 2008;58: Hoffman HM, Throne ML, Amar NJ, Sebai M, Kivitz AJ, Kavanaugh A, et al. Efficacy and safety of rilonacept (interleukin-1 Trap) in patients with cryopyrin-associated periodic syndromes: results from two sequential placebo-controlled studies. Arthritis Rheum 2008;58: