Interleukin 18 As a Marker of Disease Activity and Severity in Patients with Juvenile Idiopathic Arthritis ANA PAOLA N

Home , Interleukin 6

Interleukin 18 as a Marker of Disease Activity and Severity in Patients with Juvenile Idiopathic Arthritis ANA PAOLA N. LOTITO, ANA CAMPA, CLOVIS A.A. SILVA, MARIA H.B. KISS, and SUZANA B.V. MELLO

ABSTRACT. Objective. To verify the importance of interleukin 18 (IL-18) in the pathogenesis of juvenile idiopathic arthritis (JIA). We measured IL-18 levels in synovial fluid (SF) and serum, and determined their correlation with measures of disease activity and severity. Methods. Fifty patients with JIA (13 systemic, 13 polyarticular, 24 oligoarticular) and 25 matched controls were analyzed. Cytokine levels (IL-1ß, IL-1Ra, IL-6, and IL-18) were quantified in serum and SF by ELISA, and disease activity measures were evaluated immediately after knee articular puncture. Radiological assessment was made according to the Steinbrocker method. Statistical analysis was performed by Spearman’s rank-order correlation and Mann-Whitney rank test. Results. All the analyzed cytokine levels (IL-1, IL-1Ra, IL-6, and IL-18) were higher in patients’ sera than in controls. Remarkably, in patients with JIA, IL-18 SF levels did not differ from those of serum; they were positively correlated. The levels of IL-18 (SF and serum) were positively correlated with measures of disease activity: C-reactive protein, number of active joints, and radiological score, as well as with levels of IL-1, IL-1Ra, and IL-6. Moreover, IL-18 and IL-6 levels in SF and serum were much higher in patients with systemic disease compared to the other types of disease onset. In contrast, IL-1 and IL-1Ra were not different among JIA subtypes. Conclusion. Our results strongly suggest the participation of IL-18 in the pathophysiology of JIA. The positive correlation of this cytokine with several measures of articular inflammation and disease severity suggests that IL-18 could be a better target for the treatment of arthritis. (First Release Mar 1 2007; J Rheumatol 2007;34:823–30)

Key Indexing Terms: JUVENILE IDIOPATHIC ARTHRITIS SYNOVIAL FLUID INTERLEUKIN 18 CYTOKINES

Juvenile idiopathic arthritis (JIA) is a chronic inflammatory tion and perpetuation of inflammatory response and joint disorder of unknown etiology primarily involving joints, con- damage in JIA3, sometimes correlating with disease activity stituting the most common arthritic condition in childhood1. measures and the degree of articular damage4-8. One of the pathologic traits of the disease is the tumor-like Recent studies have revealed a broad spectrum of effects expansion of inflamed synovial tissue and infiltration of that implicate IL-18, a member of the IL-1 superfamily, as a inflammatory cells, including macrophages, plasma cells, and pivotal regulator of chronic inflammation in human autoim- lymphocytes; the production of cytokines by these cells great- mune diseases9. This cytokine, originally described as inter- ly influences the outcome of the disease2. feron-γ inducing factor (IGIF), seems to exert powerful Th1- Inflammatory cytokines secreted by activated macro- promoting activities in synergy with IL-1210. Macrophages, phages and T cells [interleukin 1 (IL-1), IL-6, and tumor Kupfer cells, keratinocytes, articular chondrocytes, synovio- necrosis factor-α (TNF-α)] are directly involved in the initia- cytes, and osteoblasts are able to synthesize IL-1811-16. Dai, et al17 showed that IL-18 augments monocyte activation by direct cell-cell contact with activated T cells in rheumatoid From the Rheumatology Division, Department of Internal Medicine, and Pediatric Department, School of Medicine, University of Sao Paulo, Sao arthritis (RA) synovitis, and dose-dependently enhances the Paulo, Brazil. production of IL-1ß and TNF-α, which promotes IL-8-medi- Supported by Fundação de Amparo à Pesquisa do Estado de São Paulo ated neutrophil chemotaxis18. Moreover, IL-18 induces the FAPESP #01/01663-6. expression of CXC chemokines by synovial fibroblasts, stim- A.P.N. Lotito, MD, PhD, Medical Assistant; C.A.A. Silva, MD, PhD, Head, Pediatric Rheumatology Unit, Department of Pediatrics; M.H.B. ulates angiogenesis, and is involved in leukocyte recruitment Kiss, MD, PhD, Professor of Pediatric Rheumatology; A. Campa, PhD, by upregulating the adhesion molecules vascular cell adhesion Professor, Department of Clinical and Toxicological Analysis, School of molecule-1 (VCAM-1) and intercellular CAM-117. Recently, Pharmaceutical Sciences; S.B.V. Mello, PhD, Associate Professor, Rheumatology Division, Department of Internal Medicine, University of it has been shown that IL-18 itself can stimulate Th2 cytokine Sao Paulo. production as well as allergic inflammation19,20. Address reprint requests to Prof. S.B.V. Mello, Rheumatology Division, Several models of arthritis have targeted IL-18. Blockade Department of Internal Medicine, School of Medicine, University of Sao Paulo, Av. Dr. Arnaldo 455, São Paulo 0124-6903 SP, Brazil. of IL-18 reduced the incidence and severity of arthritis, also 21-25 E-mail: [email protected] reducing the production of proinflammatory cytokines . Accepted for publication December 15, 2006. Studies with RA synovial cell culture have shown increased

Lotito, et al: IL-18 in JIA 823

Downloaded on September 27, 2021 from www.jrheum.org production of nitric oxide as well as TNF-α, granulocyte protein (CRP), platelet count, RF, and ANA. Knee radiographs were also 33 macrophage-colony stimulating factor, interferon-γ, and IL-6 obtained and analyzed according to the Steinbrocker method . The clinical under direct stimulation by IL-1811,26. These findings suggest data were obtained by review of the medical records of all patients and includ- α ed disease duration, duration of arthritis in the affected knee, and duration of that IL-18 is a pre-TNF- inflammatory cytokine. clinical and laboratory activity. In adults with RA, but not in patients with osteoarthritis Systemic features were defined mainly as the presence of fever and/or (OA), elevated levels of IL-18 have been described in the syn- rash, sometimes accompanied by lymphadenopathy, enlargement of liver or ovium11,26. Serum levels of IL-18 were extremely elevated in spleen, or serositis. When the disease course was analyzed, the patients with patients with active adult-onset Still’s disease (AOSD) com- extended oligoarthritis (more than 4 affected joints after 6 mo of disease) were included in the group of patients with polyarticular or systemic onset pared to the patients with RA, and correlated with some meas- who still had 5 or more affected joints with no systemic manifestation after 6 ures of disease activity27,28. months of disease duration. There is one study evaluating serum levels of IL-18 in SF analysis and quantification of cytokines. Immediately after the articular patients with JIA showing extremely high levels of this puncture, total cells were counted in a Newbauer chamber and differential cytokine only in patients with systemic disease compared to counts were performed in smears stained with Giemsa solution. SF and serum other types of onset29. Heinzmann, et al genotyped 5 IL-18 were immediately stored at –80°C for cytokine determination. At the time of the assay diluted serum and SF IL-1, IL-1 receptor antagonist (IL-1Ra), IL-6 polymorphisms in patients with antinuclear antibody (ANA)- (R&D Systems Inc., Minneapolis, MN, USA), and IL-18 levels (Medical and positive JIA, but did not find any association with the dis- Biological Laboratories Co., Nagoya, Japan) were evaluated by commercial ease30. Scola, et al evaluated the synovium of patients with ELISA according to the manufacturer’s instructions. The minimum detectable polyarticular JIA and analyzed cytokines supposed to be concentrations of the cytokines were as follows: IL-1ß (0.1 pg/ml), IL-1Ra important for the priming and perpetuation of type 1 respons- (40 pg/m), IL-6 (15 pg/ml), and IL-18 (25 pg/ml). es, showing that IL-12, IL-15, and IL-18 are widely expressed Statistical analysis. Results are expressed as median (range). Within-group 31 comparisons were analyzed by Student’s t test (when normal distribution was in highly infiltrated tissue sections . detected) or Mann-Whitney rank-sum test. Correlations were expressed using These data suggest that IL-18 is also involved in the patho- the Spearman rank correlation coefficient. For the analyzed measures, p val- genesis of JIA; however, until now there have been no studies ues less than 0.05 were considered significant. with children analyzing this new cytokine in the synovial fluid (SF). RESULTS Our aim was to evaluate the importance of IL-18 in JIA Clinical features. Results depicted in Table 1 express the main pathogenesis, and establish possible correlations between this clinical and laboratory characteristics of the study population cytokine and measures of disease activity/articular damage, as in general and divided into groups based on the types of dis- well as with the levels of other inflammatory cytokines in SF ease onset. It is important to emphasize that patients with and serum. oligoarticular onset had a shorter disease duration (vs polyarticular, p = 0.001; vs systemic, p = 0.008), fewer joints MATERIALS AND METHODS with limited range of motion (vs polyarticular, p = 0.03; vs Patients. From 1999 to 2003, 50 consecutive patients with the diagnosis of systemic, p = 0.04), and lower CRP levels than patients with 1 JIA (ILAR criteria) (38 female) who underwent knee steroid injection and 25 polyarticular or systemic onset (vs polyarticular, p = 0.006; vs age- and sex-matched healthy controls were included in the study. The mean age of the patients was 11.6 years (range 3.6–22 yrs) and mean disease dura- systemic, p = 0.0002). Patients with systemic onset had a tion was 6.9 years (range 0.2–20 yrs). Regarding the type of onset, 13 patients higher number of active joints (p = 0.02) and a higher SF were systemic, 13 polyarticular [2 rheumatoid factor (RF)-positive], and 24 leukocyte count than other types of disease onset (vs oligo, p oligoarticular. = 0.03; vs polyarticular, p = 0.012). The majority of patients The control population consisted of 25 healthy children and adolescents in the 3 groups had no or minor functional disability, although seen in the pediatric clinic, with mean age of 10.4 years (range 1.6–20 yrs). The appropriate institutional ethics review committee approved the study most of them had advanced radiological scores (class III or and written informed consent was obtained from each patient or their parents, IV), except for patients with oligoarticular disease. All before enrolment in the study. Permission to draw extra blood during routine patients were taking NSAID; the majority of patients with pol- venipuncture was obtained from the parents of all children. yarticular onset were also receiving DMARD, and in the sys- All patients were taking nonsteroidal antiinflammatory drugs (NSAID); in addition, approximately 2/3 of them were using disease modifying temic group most were receiving triple therapy (NSAID, antirheumatic drugs [DMARD: methotrexate (MTX), n = 15, sulfasalazine, n DMARD, and steroids). = 11, chloroquine, n = 3, and cyclosporine, n = 5]. The mean dose of MTX IL-18 in patients with JIA. Patients with JIA presented serum was 13.7 mg/m2/week and 15 patients were taking oral prednisone (mean dose 0.3 mg/kg/day; range 2.5–20 mg). levels of IL-18 higher than those of controls, 590 (range All children were clinically evaluated just before knee articular puncture. 88–4000) vs 398 (76–760) pg/ml, respectively (p = 0.002; The number of active joints (swelling or limitation of joint movement with Table 2). Although IL-18 serum and SF levels did not differ, joint pain or tenderness), number of joints with limited range of motion, dura- we found a positive correlation between them (r = 0.68; p < tion of morning stiffness, functional status (American College of 0.0001). Rheumatology criteria)32, physician global evaluation of disease activity, and the parental assessment of child’s overall well-being (both measured by a 10 In SF and sera, IL-18 levels were markedly increased in cm visual analog scale) were also recorded. The following laboratory tests systemic patients compared to the other types of disease onset were performed at this visit: erythrocyte sedimentation rate (ESR), C-reactive (p < 0.0001; Figure 1). However, we did not verify any dif-

824 The Journal of Rheumatology 2007; 34:4

Downloaded on September 27, 2021 from www.jrheum.org Table 1. Clinical and demographic data of patients with juvenile idiopathic arthritis (JIA).

JIA, Oligoarticular Polyarticular Systemic Onset, n = 50 Onset, n = 24 Onset, n = 13 n = 13

Mean age, yrs 11.6 (3.6–22) 9.8 (3.6–16.4) 13.4 (6.7–21) 13 (4–21) Sex, male:female 1:3.1 1:5 1:4.3 1:2.6 Disease duration, yrs* (range) 6.3 (0.2–20) 2.7 (0.2–13) 8 (4–20) 4.7 (1.2–16) Number of active joints* (range) 2 (1–25) 2 (1–5) 2 (2–4) 2 (2–25) Number of joints with LRM* (range) 5 (1–50) 2 (1–25) 2.5 (1–50) 2.5 (3–33) ACR functional status, n (%) I+II, 41 (82) I+II, 20 (84) I+II, 12 (92) I+II, 9 (70) III + IV, 9 (18) III+IV, 4 (16) III+IV, 1 (8) III+IV, 4 (30) CRP, mg/l* (range) 13 (0–193) 1 (0–69) 12 (0–145) 24 (0–193) ESR, mm/h* (range) 37 (6–82) 32 (6–61) 40 (13–55) 36 (8–82) Number of SF leukocytes, 17,800 17,300 3050 18,600 cells/mm3* (range) (0–112,000) (0–91,000) (400–112,000) (2500–102,000) RF-positive, n (%) 4 (8) 2 (8) 2 (15) 0 Radiological score, n (%) III+II 20 (40) I+II, 13 (54) I+II, 5 (38) I+II, 2 (15) III+IV, 30 (60) III+IV, 11 (46) III+IV, 8 (62) III+IV, 11 (85) Physician global evaluation of disease 4.2 (2–7) 3.8 (2–7) 4.5 (3–7) 4.6 (3–7) activity, 10 cm VAS NSAID use, n (%) 17 (34) 15 (62) 3 (23) 0 NSAID + DMARD 18 (36) 6 (25) 9 (69) 3 (23) NSAID + DMARD + steroid 15 (30) 3 (12) 2 (15) 10 (77)

* Median (range). LRM: limited range of motion; ACR: American College of Rheumatology; CRP: C-reactive protein; ESR: erythrocyte sedimentation rate; RF: rheumatoid factor; SF: synovial fluid; NSAID: nonsteroidal antiinflammatory drug; DMARD: disease modifying antirheumatic drug.

Table 2. Inflammatory cytokines (IL-1, IL-1Ra, IL-6, and IL-18) in serum and SF of patients with JIA and controls. Data are median (range).

Cytokine Controls, JIA, Oligoarticular, Polyarticular, Systemic, n = 25 n = 50 n = 24 n = 13 n = 13

Serum IL-18 398 (76–760) 590 (88–4000)* 540 (88–920)*† 430 (300–920)*† 1760 (756–4000)* Serum IL-1 7.3 (6.4–29.9) 9.9 (6.8–389)* 11.3 (6.8–30.8)* 9 (7.32–88.2)* 12.6 (8.06–389)* Serum IL-1Ra 355 (284–546) 546 (151–11,871)* 546 (364–1705)* 453 (315–1523)* 586 (151–11,871)* Serum IL-6 7.1 (6–10.9) 11.4 (4.2–470)* 9.19 (4.2–90)*† 12.9 (6–260)* 39.5 (4.3–470)* SF IL-18 — 860 (70–2800) 750 (70–1597)† 720 (86–960)† 1709 (410–2800) SF IL-1 — 18.4 (6.4–350) 22 (6.6–2825) 14 (6.4–152.4) 17.3 (6.9–350.3) SF IL-1Ra — 4126 (797–30,903) 4126 (829–30,903) 2305 (797–22,272) 5883 (1217–29,890) SF IL-6 — 1600 (30–3620) 1680 (64–3280)† 720 (30–3000)† 2720 (235–3600)

* p < 0.05 versus control; † p < 0.05 versus systemic. ference between oligoarticular and polyarticular onset IL-18 and measures of disease activity and damage. patients. Analyzing the patients as a single group, IL-18 correlated with As the majority of patients with systemic onset disease several measures of disease activity and damage (Table 3). evolved to polyarticular course, and only 5 patients still had Both in serum and in SF, IL-18 levels positively correlated systemic manifestations when the study was developed, we with CRP, number of active joints, and radiological score. compared IL-18 levels according to the type of disease pro- Synovial IL-18 also correlated with the number of infiltrating gression, and observed that patients with systemic course (n = leukocytes (rs = 0.46; p = 0.0001) and polymorphonuclear 5) still had higher levels of IL-18 than the patients with pol- cells (r = 0.43; p = 0.002). yarticular (n = 22) or oligoarticular course (n = 22), both in No correlation was found between IL-18 levels and func- serum and in SF (p < 0.05). tional status, parental assessment of child’s overall well- Of note, at the time of knee steroid injection the levels of being, disease duration, duration of arthritis in the affected IL-18 in the SF were similar in the patients with systemic JIA knee, duration of clinical and laboratory activity, morning with or without systemic features (median 1720 and 1680 stiffness, or RF. pg/ml, respectively). Similar findings were detected in serum In patients with polyarticular onset, SF IL-18 expressed a (median 1800 and 1680 pg/ml). high correlation with physician global evaluation of disease

Lotito, et al: IL-18 in JIA 825

Downloaded on September 27, 2021 from www.jrheum.org Figure 1. IL-18 levels in serum (A) and synovial fluid (panel B) of 50 patients with JIA subdivided according to type of disease onset, and controls. Bars represent median of values, , * represent extreme values. Systemic disease vs control, p = 0.0001; polyarticular vs control, p = 0.04; oligoarticular vs control, p = 0.02. B. Systemic vs polyarticular, p = 0.0001; systemic vs oligoarticular, p = 0.0001. activity (by visual analog scale: rs = 0.77; p = 0.001). No other = 17), patients taking NSAID and DMARD (n = 18), and correlations were found between the analyzed measures and patients taking NSAID and DMARD plus oral steroids (n = IL-18 when patients were subdivided according to the types of 15). Interestingly, patients receiving corticosteroids had high- disease onset. er SF levels of IL-18 than patients taking DMARD (p = 0.003) IL-18 and treatment. In order to evaluate whether the treat- or only NSAID (p = 0.0002). The same occurred with serum ment interferes with serum and SF levels of IL-18, the patients levels of IL-18 (p < 0.05). were divided into 3 groups: patients receiving only NSAID (n When we compared clinical and laboratory data among

826 The Journal of Rheumatology 2007; 34:4

Downloaded on September 27, 2021 from www.jrheum.org Table 3. Correlations between serum or SF IL-18 levels and measures of serum of patients with JIA compared to controls. We also disease activity/damage and with inflammatory cytokines. observed that these cytokine levels were enhanced in SF compared to serum (p < 0.05; Table 2). Additionally, IL-18 levels Synovial Fluid IL-18 Serum IL-18 r pr p in serum and SF positively correlated with the levels of IL-1, s s IL-1Ra, and IL-6 (Table 2, Figure 2). Number of active joints 0.39 0.03 0.50 < 0.0001 IL-1 and IL-1Ra levels were similar in JIA onset types. C-reactive protein 0.48 < 0.0001 0.47 < 0.0001 Nevertheless, SF IL-6 levels in systemic patients were higher Radiological score 0.42 0.003 0.37 0.008 than in polyarticular (p < 0.0001) and oligoarticular onset Infiltrating leukocytes 0.46 < 0.0001 — IL-1 0.42 0.003 0.31 0.04 patients (p = 0.007). Enhanced serum levels of IL-6 were IL-1Ra 0.42 0.002 0.34 0.025 observed in the group with systemic onset compared to the IL-6 0.33 0.02 0.48 < 0.0001 patients with oligoarticular onset disease (p = 0.002). The levels of IL-1 positively correlated with the levels of IL-1Ra in serum (r = 0.98; p < 0.0001) and SF (r = 0.70; p < these 3 treatment groups, we observed that patients taking 0.0001) and did not correlate with measures of disease activi- steroids had a longer disease duration (p = 0.005) and a more ty, except with polymorphonuclear cells in SF (rs = 0.29, p = severe disease than patients taking NSAID, with worse radio- 0.04). logical scores (p = 0.0003), higher number of joints with lim- Besides the correlation with IL-18, IL-6 SF levels also cor- ited range of motion (p < 0.0001), higher CRP (p = 0.006), related with IL-1 (rs = 0.42; p = 0.004) and IL-1Ra (rs = 0.39; and higher number of infiltrating leukocytes (p = 0.002). p = 0.007). It was interesting to observe that the number of IL-18 and other proinflammatory cytokines. To investigate the infiltrating leukocytes and polymorphonuclear cells was the association between IL-18 and other cytokines known to be only inflammatory measure associated with synovial levels of involved in JIA pathogenesis, levels of IL-1, IL-1Ra, and IL- IL-6 (rs = 0.47; p < 0.0001 and rs = 0.49; p < 0.0001). On the 6 were quantified in patients and controls. other hand, serum IL-6 correlated with several measures of Higher levels of IL-1, IL-1Ra, and IL-6 were observed in disease activity: CRP (rs = 0.56; p < 0.0001), ESR (r = 0.41; p

Figure 2. Positive correlations between IL-18 levels measured in 50 patients with JIA and IL-1 (A, synovial fluid; B, serum) and IL-6 (C, synovial fluid; D, serum).

Lotito, et al: IL-18 in JIA 827

Downloaded on September 27, 2021 from www.jrheum.org = 004); number of active joints (rs = 0.70; p < 0.0001), num- complication of systemic JIA, is associated with a high number of joints with limited range of motion (rs = 0.39; p = ber of cells expressing IL-18 in the bone marrow but not in 0.006), radiological score (rs = 0.30; p = 0.006), and number other organs, suggesting that bone marrow is the origin of 36 of platelets (rs = 0.50; p < 0.0001). increased IL-18-producing cells that occur in this disease . In adults, IL-18 seems to be an excellent marker of disease DISCUSSION activity, showing frequent correlation with levels of acute Our study showed, for the first time, a positive correlation phase reactants11,27,29. Our results showed that both serum between serum and intraarticular levels of IL-18 in patients and SF IL-18 levels correlated with CRP, number of swollen with JIA. Further, our results showed elevated serum levels of joints, and inflammatory cytokines. One study indicated that this cytokine in patients with JIA compared to the control serum levels of IL-18 correlated well with clinical activity group. Our data are in agreement with another study carried score and predicted the occurrence of liver dysfunction35. out with adults, which found that the amount of IL-18 present Another study has found a positive correlation between IL-18 in serum and SF of patients with RA was markedly increased serum levels and serum ferritin, but not with CRP levels29. On when compared to patients with OA11. the other hand, previous authors found no association with IL-18-binding protein (IL-18BP), a constitutively secreted measures of disease activity28,37. protein that binds mature IL-18 with high affinity, provides a In our study, IL-18 levels in serum and SF correlated with potential mechanism to regulate the activity of this cytokine34. radiological scores. Gracie, et al showed an increase in the The ratio between IL-18 and its inhibitors is responsible for erosive and inflammatory component of mice with collagen- the bioactivity of this cytokine. Kawashima, et al found that induced arthritis injected with IL-1826. Another group showed IL-18BP mRNA expression in whole blood of patients with attenuation of the disease after blockade of IL-18 demonstrat- RA was lower than in controls; this may enhance IL-18 bioac- ed by reduced evidence of cartilage erosion24. It seems that tivity, leading to the proinflammatory profile characteristic of this cytokine is involved in the catabolism of articular carti- this disease27. In our study, the levels of IL-18 did not differ lage by upregulating stromelysin and increasing proteoglycan between serum and SF, which does not reflect the presence of release14. inhibitors such as IL-18BP in the synovial compartment, since Joosten, et al found that biopsy specimens of knee synovial this does not interfere with the immunoassay used. membrane expressed IL-18 in 80% of the patients with RA38. Although all our patients presented higher levels of IL-18 IL-18 expression in the synovial tissue was strongly correlat- than the control group, we observed that systemic patients had ed with the expression of IL-1ß and TNF-α, macrophage infil- serum and SF levels that were markedly elevated compared to tration, microscopic inflammation scores, and ESR. In our those in other types of disease onset. These patients were the study, synovial IL-18 was strongly correlated with the number ones with the most severe inflammatory process, expressed by of infiltrating leukocytes, suggesting that this cytokine can the highest number of active joints, CRP levels, and leukocyte actually be involved in the recruitment of inflammatory cells count in the SF. IL-18 has been shown to be involved in the to the SF. pathogenesis of arthritis with systemic manifestations, both in It is important to emphasize that we observed that patients children and adults. Chen, et al found significantly higher lev- taking corticosteroids in association with DMARD or NSAID els of IL-6, IL-8, IL-18, and TNF-α in sera from patients with had higher levels of IL-18 in serum and SF. This probably active untreated AOSD compared to healthy controls, and also reflects a greater inflammatory process and disease severity found elevated levels of IL-18 transcripts in the biopsy speci- (these patients had a worse radiological score, more affected mens of Still’s rash and synovitis35. Kawashima, et al showed joints, higher CRP, and number of infiltrating leukocytes). that in patients with AOSD, the levels of IL-18 became Kawaguchi, et al observed that patients with AOSD who were markedly elevated when pyrexia and arthralgia developed, responsive to corticosteroid treatment had lower levels of IL- and correlated with serum ferritin values and disease severi- 18 than nonresponsive patients, suggesting that this cytokine ty27. Similarly to our study, Maeno, et al studied children with is a specific indicator of severity in AOSD28. JIA, observing extremely high serum levels of this cytokine We also observed that IL-18 levels in serum and SF were only in patients with systemic onset, without significant dif- positively correlated with the levels of IL-6 and were also ferences among patients with polyarticular JIA and oligoartic- enhanced in JIA sera compared to the control group. IL-6 is ular JIA and healthy children29. These authors found that elevated in patients with JIA in general, but plays a major role within the systemic JIA group, patients with serositis or in the pathogenesis of the systemic onset disease type5-7,35,39-42. hepatosplenomegaly had significantly higher levels of IL-18 Other authors have shown that IL-18 is able to induce the pro- than patients without such manifestations. Within our group of duction of IL-6 both in experimental arthritis and in normal patients, only 5 had a systemic disease course. Therefore, any human articular chondrocytes14,22, and that neutralization of association between IL-18 levels and systemic features is dif- the former significantly reduced serum levels of IL-624. ficult to establish. It is important to emphasize that IL-6 may not be the best The macrophage activation syndrome, the most severe disease activity marker when measured in the SF, since it cor-

828 The Journal of Rheumatology 2007; 34:4

Downloaded on September 27, 2021 from www.jrheum.org related only with the number of infiltrating leukocytes. 9. Dinarello CA. Interleukin-18, a proinflammatory cytokine. Eur Nevertheless, IL-18 also correlated with several variables of Cytokine Netw 2000;11:483-6. inflammation and disease severity when measured in SF as 10. Ushio S, Namba M, Okura T, et al. Cloning of the cDNA for human IFN-gamma-inducing factor, expression in Escherichia coli, and well as in serum (number of active joints, CRP, radiological studies on the biologic activities of the protein. J Immunol score, levels of inflammatory cytokines, number of infiltrating 1996;156:4274-9. leukocytes, physician evaluation of disease activity), suggest- 11. Yamamura M, Kawashima M, Taniai M, et al. ing that this cytokine is involved in joint damage and may be Interferon-gamma-inducing activity of interleukin-18 in the joint an effective therapeutic approach. with rheumatoid arthritis. Arthritis Rheum 2001;44:275-85. 12. Kohno K, Kurimoto M. Interleukin 18, a cytokine which resembles There is controversy about IL-1 and IL-1Ra levels in IL-1 structurally and IL-12 functionally but exerts its effect patients with JIA. Studies have demonstrated elevated and independently of both. Clin Immunol Immunopathol 1998;86:11-5. normal levels of this cytokine both in serum and in SF43,44. 13. Udagawa N, Horwood NJ, Elliott J, et al. Interleukin-18 The correlation with disease activity is also variable42,44,45. (interferon-gamma-inducing factor) is produced by osteoblasts and IL-1Ra has been found to be correlated with the fever peak in acts via granulocyte/macrophage colony-stimulating factor and not via interferon-gamma to inhibit osteoclast formation. J Exp Med patients with systemic JIA, but has not differed among the 1997;185:1005-12. onset types46. We found a positive correlation between serum 14. Olee T, Hashimoto S, Quach J, Lotz M. IL-18 is produced and SF IL-18 and levels of IL-1 and IL-1Ra; in addition, lev- by articular chondrocytes and induces proinflammatory and els of both cytokines were found to be higher in SF than in catabolic responses. J Immunol 1999;162:1096-100. serum, suggesting local production. 15. Moller B, Kessler U, Rehart S, et al. Expression of interleukin-18 receptor in fibroblast-like synoviocytes. Arthritis Res Our results showed that IL-18 is a good marker of disease 2002;4:139-44. activity and damage when measured in serum and SF, sug- 16. Takeda K, Tsutsui H, Yoshimoto T, et al. Defective NK cell activity gesting the importance of this cytokine in the pathogenesis of and Th1 response in IL-18-deficient mice. Immunity JIA. These data situate IL-18 in a key position in the complex 1998;8:383-90. network of chemical mediation of joint inflammation, and as 17. Dai SM, Nishioka K, Yudoh K. Interleukin (IL) 18 stimulates osteoclast formation through synovial T cells in rheumatoid a new marker and possibly a better target for JIA therapy. arthritis: comparison with IL-1 beta and tumour necrosis factor alpha. Ann Rheum Dis 2004;63:1379-86. ACKNOWLEDGMENT 18. Puren AJ, Fantuzzi G, Gu Y, Su MS, Dinarello CA. Interleukin-18 We thank Dr. Hilario Salomão Joffe for helping with the selection of the con- (IFN gamma-inducing factor) induces IL-8 and IL-1 beta via TNF trol group, Drs. Eloisa Bonfa and Gilberto Santos Novaes for helpful discus- alpha production from non-CD14+ human blood mononuclear sion, and Maria Aurora Gomes and Maria de Fátima de Almeida for technical cells. J Clin Invest 1998;101:711-21. assistance. 19. Nakanishi K, Tsutsui H, Yoshimoto T. IL-18 and atopic inflammation. Tanpakushitsu Kakusan Koso 2002;47 Suppl:2362-8. REFERENCES 20. Hoshino T, Yagita H, Ortaldo JR, Wiltrout RH, Young HA. In vivo administration of IL-18 can induce IgE production through Th2 1. Petty RE, Southwood TR, Manners P, et al; International League of cytokine induction and up-regulation of CD40 ligand (CD154) Associations for Rheumatology. International League of expression on CD4+ T cells. Eur J Immunol 2000;30:1998-2006. Associations for Rheumatology classification of juvenile idiopathic 21. Wei XQ, Leung BP, Arthur HM, McInnes IB, Liew FY. Reduced arthritis: second revision, Edmonton, 2001. J Rheumatol incidence and severity of collagen-induced arthritis in mice lacking 2004;31:390-2. IL-18. J Immunol 2001;166:517-21. 2. Wynne-Roberts CR, Anderson CH, Turano AM, Baron M. 22. Leung BP, McInnes IB, Esfandiari E, Wei XQ, Liew FY. Combined Light- and electron-microscopic findings of juvenile rheumatoid effects of IL-12 and IL-18 on the induction of collagen-induced arthritis synovium: comparison with normal juvenile synovium. arthritis. J Immunol 2000;164:6495-502. Semin Arthritis Rheum 1978;7:287-302. 23. Lubberts E, van den Berg WB. Cytokines in the pathogenesis of 3. Woo P. Cytokines in childhood rheumatic diseases. Arch Dis Child rheumatoid arthritis and collagen-induced arthritis. Adv Exp Med 1993;69:547-9. Biol 2003;520:194-202. 4. Eberhard BA, Laxer RM, Andersson U, Silverman ED. Local 24. Plater-Zyberk C, Joosten LA, Helsen MM, et al. Therapeutic effect synthesis of both macrophage and T cell cytokines by synovial fluid cells from children with juvenile rheumatoid arthritis. Clin of neutralizing endogenous IL-18 activity in the collagen-induced Exp Immunol 1994;96:260-6. model of arthritis. J Clin Invest 2001;108:1825-32. 5. Rooney M, David J, Symons J, Di Giovine F, Varsani H, Woo P. 25. Smeets RL, van de Loo FA, Arntz OJ, Bennink MB, Joosten LA, Inflammatory cytokine responses in juvenile chronic arthritis. Br van den Berg WB. Adenoviral delivery of IL-18 binding protein C J Rheumatol 1995;34:454-60. ameliorates collagen-induced arthritis in mice. Gene Ther 6. De Benedetti F, Robbioni P, Massa M, Viola S, Albani S, Martini A. 2003;10:1004-11. Serum interleukin-6 levels and joint involvement in polyarticular 26. Gracie JA, Forsey RJ, Chan WL, et al. A proinflammatory role for and pauciarticular juvenile chronic arthritis. Clin Exp Rheumatol IL-18 in rheumatoid arthritis. Clin Invest 1999;104:1393-401. 1992;10:493-8. 27. Kawashima M, Yamamura M, Taniai M, et al. Levels of 7. De Benedetti F, Martini A. Is systemic juvenile rheumatoid arthritis interleukin-18 and its binding inhibitors in the blood circulation of an interleukin 6 mediated disease? J Rheumatol 1998;25:203-7. patients with adult-onset Still’s disease. Arthritis Rheum 8. Gattorno M, Facchetti P, Ghiotto F, et al. Synovial fluid T cell 2001;44:550-60. clones from oligoarticular juvenile arthritis patients display a 28. Kawaguchi Y, Terajima H, Harigai M, Hara M, Kamatani N. prevalent Th1/Th0-type pattern of cytokine secretion irrespective of Interleukin-18 as a novel diagnostic marker and indicator of disease immunophenotype. Clin Exp Immunol 1997;109:4-11. severity in adult-onset Still’s disease. Arthritis Rheum

Lotito, et al: IL-18 in JIA 829

Downloaded on September 27, 2021 from www.jrheum.org 2001;44:1716-7. 38. Joosten LA, Radstake TR, Lubberts E, et al. Association of 29. Maeno N, Takei S, Nomura Y, Imanaka H, Hokonohara M, Miyata interleukin-18 expression with enhanced levels of both K. Highly elevated serum levels of interleukin-18 in systemic interleukin-1 beta and tumor necrosis factor alpha in knee synovial juvenile idiopathic arthritis but not in other juvenile idiopathic tissue of patients with rheumatoid arthritis. Arthritis Rheum arthritis subtypes or in Kawasaki disease: comment on the article 2003;48:339-47. by Kawashima et al. Arthritis Rheum 2002;46:2539-41; author 39. Hirano T. Cytokines in autoimmune disease and chronic reply 2541-2. inflammatory proliferative disease. Cytokine Growth Factor Rev 30. Heinzmann A, Gerhold K, Ganter K, et al. Association study of 2002;13:297-8. polymorphisms within interleukin-18 in juvenile idiopathic arthritis 40. Lepore L, Pennesi M, Saletta S, Perticarari S, Presani G, Prodan M. and bronchial asthma. Allergy 2004;59:845-9. Study of IL-2, IL-6, TNF alpha, IFN gamma and beta in the serum 31. Scola MP, Thompson SD, Brunner HI, et al. Interferon- and synovial fluid of patients with juvenile chronic arthritis. Clin gamma:interleukin 4 ratios and associated type 1 cytokine Exp Rheumatol 1994;12:561-5. expression in juvenile rheumatoid arthritis synovial tissue. 41. Prieur AM, Roux-Lombard P, Dayer JM. Dynamics of fever and the J Rheumatol 2002;29:369-78. cytokine network in systemic juvenile arthritis. Rev Rhum Engl Ed 32. Hochberg MC, Chang RW, Dwosh I, Lindsey S, Pincus T, Wolfe F. 1996;63:163-70. The American College of Rheumatology 1991 revised criteria for 42. Mangge H, Kenzian H, Gallistl S, et al. Serum cytokines in juvenile the classification of global functional status in rheumatoid arthritis. rheumatoid arthritis. Correlation with conventional inflammation Arthritis Rheum 1992;35:498-502. parameters and clinical subtypes. Arthritis Rheum 1995;38:211-20. 33. Steinbrocker O, Traeger CH, Batterman RC. Therapeutic criteria in 43. Martini A, Ravelli A, Notarangelo LD, et al. Enhanced interleukin 1 rheumatoid arthritis. JAMA 1949;140:659-62. and depressed interleukin 2 production in juvenile arthritis. 34. Novick D, Schwartsburd B, Pinkus R, et al. A novel IL-18BP J Rheumatol 1986;13:598-603. ELISA shows elevated serum IL-18BP in sepsis and extensive 44. Ou LS, See LC, Wu CJ, Kao CC, Lin YL, Huang JL. Association decrease of free IL-18. Cytokine 2001;14:334-42. between serum inflammatory cytokines and disease activity in 35. Chen DY, Lan JL, Lin FJ, Hsieh TY. Proinflammatory cytokine juvenile idiopathic arthritis. Clin Rheumatol 2002;21:52-6. profiles in sera and pathological tissues of patients with active 45. Kutukculer N, Caglayan S, Aydogdu F. Study of pro-inflammatory untreated adult onset Still’s disease. J Rheumatol 2004;31:2189-98. (TNF-alpha, IL-1 alpha, IL-6) and T-cell-derived (IL-2, IL-4) 36. Maeno N, Takei S, Imanaka H, et al. Increased interleukin-18 cytokines in plasma and synovial fluid of patients with juvenile expression in bone marrow of a patient with systemic juvenile chronic arthritis: correlations with clinical and laboratory idiopathic arthritis and unrecognized macrophage-activation parameters. Clin Rheumatol 1998;17:288-92. syndrome. Arthritis Rheum 2004;50:1935-8. 46. De Benedetti F, Pignatti P, Massa M, Sartirana P, Ravelli A, Martini 37. Bresnihan B, Roux-Lombard P, Murphy E, Kane D, FitzGerald O, A. Circulating levels of interleukin 1 beta and of interleukin 1 Dayer JM. Serum interleukin 18 and interleukin 18 binding protein receptor antagonist in systemic juvenile chronic arthritis. Clin Exp in rheumatoid arthritis. Ann Rheum Dis 2002;61:726-9. Rheumatol 1995;13:779-84.

830 The Journal of Rheumatology 2007; 34:4

Downloaded on September 27, 2021 from www.jrheum.org