- Email: [email protected]
Identification of the best cutoff points and clinical signs specific for early recognition of macrophage activation syndrome in active systemic juvenile idiopathic arthritis
Seminars in Arthritis and Rheumatism ] (2014) ]]]–]]]
Contents lists available at ScienceDirect
Seminars in Arthritis and Rheumatism journal homepage: www.elsevier.com/locate/semarthrit
Identification of the best cutoff points and clinical signs specific for early recognition of macrophage activation syndrome in active systemic juvenile idiopathic arthritis Mikhail M. Kostik, MD, PhDn, Margarita F. Dubko, MD, PhD, Vera V. Masalova, MD, Ludmila S. Snegireva, MD, Tatyana L. Kornishina, MD, Irina A. Chikova, MD, Tatyana S. Likhacheva, MD, Eugenia A. Isupova, MD, Natalia I. Glebova, MD, Ekaterina M. Kuchinskaya, MD, Eugenia V. Balbotkina, MD, Natalia V. Buchinskaya, MD, Olga V. Kalashnikova, MD, PhD, Vyacheslav G. Chasnyk, MD, PhD Hospital Pediatry, Saint-Petersburg State Pediatric Medical University of Ministry of Healthcare of the Russia, 194100, Litovskaya 2, Saint-Petersburg, Russia
a r t i c l e in fo
Keywords: Systemic juvenile idiopathic arthritis Macrophage activation syndrome (MAS)
a b s t r a c t Objectives: The purpose of our study was to detect early clinical and laboratory signs that help to discriminate macrophage activation syndrome (MAS) from active systemic juvenile idiopathic arthritis (SJIA) without MAS. Methods: Our retrospective study was based on reviewing the medical charts of the children admitted to the rheumatology department with active SJIA and definite MAS (n ¼ 18) and without MAS (n ¼ 40). We evaluated the data related to SJIA and MAS at the moment of the patient's admission. If the patient had signs of MAS since admission or developed definite MAS later during this flare, he was referred to the main group. The children who did not have MAS during the flare episode and did not have MAS in the past medical history were in the control group. We calculated the cutoff points for MAS parameters, performed the analysis of sensitivity and specificity, identified the predictors, and provided the preliminary diagnostic rule through “the-number-of-criteria-present” approach. Results: The clinical signs were relevant to MAS in SJIA: oligoarticular disease course (OR ¼ 5.6), splenomegaly (OR ¼ 67.6), hemorrhages (OR ¼ 33.0), and respiratory failure (OR ¼ 11.3). The involvement of wrist (OR ¼ 0.2), MCP (OR ¼ 0.1), and PIP joints (OR ¼ 0.1) was protective against MAS development. The best cutoffs for laboratory parameters were PLT r 211 109/l, WBC r 9.9 109/l, AST 4 59.7 U/l, LDH 4 882 U/l, albumin r 2.9 g/dl, ferritin 4 400 μg/l, fibrinogen r 1.8 g/l, and proteinuria. The laboratory variables were more precise in the discrimination of early MAS than clinical: any 3 or more laboratory criteria provided the highest specificity (1.0) and sensitivity (1.0) and OR ¼ 2997. Conclusions: We detected clinical and laboratory markers and created preliminary diagnostic (laboratory) guidelines for early discrimination of MAS in active SJIA. & 2014 Elsevier Inc. All rights reserved.
Abbreviations: ALP, alkaline phosphatase; ALT, alanine aminotransferase; anti-IL1, anti-interleukin -1; anti-IL6, anti-interleukin-6; AST, aspartate aminotransferase; AUC, area under the curve; 95% CI, 95% confidence interval; CNS, central nervous system; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; GGTT, γ-glutamintranspeptidase; Hb, hemoglobin; HLH, hemophagocytic lymphohistiocytosis; ILAR, International League of Associations for Rheumatology; IQR, interquartile range; LDH, lactate dehydrogenase; MAS, macrophage activation syndrome; MCP, metacarpophalangeal joints; Me, median; Naþ , sodium; OR, odds ratio; PIP, proximal interphalangeous joints; PLT, platelets; SJIA, systemic juvenile idiopathic arthritis; TMJ, temporomandibular joints; TG, triglycerides; WBC, white blood cells. n Corresponding author. E-mail addresses: [email protected], [email protected] (M.M. Kostik). http://dx.doi.org/10.1016/j.semarthrit.2014.09.004 0049-0172/& 2014 Elsevier Inc. All rights reserved.
Introduction Macrophage activation syndrome (MAS) is a severe lifethreatening hematological condition, mostly complicated by systemic juvenile idiopathic arthritis (SJIA) [1]. The diagnosis of MAS can be difficult, especially in subclinical forms and in the early stage of MAS [2,3]. The early detection of MAS can lead to appropriate therapeutic interventions and change the outcomes. There are no strict criteria with enough sensitivity and specificity that are good for early MAS detection in SJIA [4]. MAS belongs to the family of histiocytic disorders, especially close to
2
M.M. Kostik et al. / Seminars in Arthritis and Rheumatism ] (2014) ]]]–]]]
hemophagocytic lymphohistiocytosis (HLH), but there are some differences in the pathogenesis between HLH and MAS that lead to different therapeutic approaches [5–10]. HLH is a typical genetic disorder related to T-cell and macrophage uncontrolled expansion linked to NK-cell and cytotoxic T-cell function considered to be the result of homozygous mutations in cytolytic pathway genes [11–16]. The pathogenesis of MAS is more complex and heterogenous. Decreased NK-cell activity, repeated stimulation, and hyperactivation of Toll-like receptor (TLR-9) and cytokine disbalance are other known pathogenic mechanisms underlying the MAS pathogenesis [17–21]. Despite the fact that MAS bears a close resemblance to hemophagocytic lymphohistiocytosis (HLH), the often applied HLH criteria created for genetic disorders are not applicable to early recognition of MAS in SJIA patients [22]. The prominent inflammation with an increased number of WBC and platelets and hyperfibrinogenemia typical for SJIA required a higher threshold of these parameters for MAS detection than in HLH. Fever, one of the main symptoms in SJIA and MAS/HLH, is very frequent in both conditions, so the sensitivity of fever is low. The main difference is the pattern of fever: non-remitting fever is typical for MAS/HLH and spiking remitting fever is more relevant for SJIA [23]. Due to the high intensity of thresholds of the HLH criteria they can only determinate the advanced stage of MAS, which leads to the delay in diagnosis and late start of specific treatment, which is associated with poor outcomes. Preliminary criteria of MAS in SJIA have been proposed, but these criteria had some limitations due to the lack of several laboratory measurements [24]. Several studies showed that clinical signs usually appear later and lead to the delay of MAS diagnosis, so the laboratory markers could be more sensitive and discriminative in early MAS diagnosis [24,25]. The purpose of our study was to detect clinical and laboratory signs that help to discriminate patients with an increased risk of MAS development or early MAS in active SJIA patients.
Methods Study design and patient selection Our retrospective study was based on reviewing the medical charts of the children who were admitted to our rheumatology department in 2005–2013 with the onset or flare of SJIA. The diagnosis of SJIA was established on ILAR definitions [26]. We collected all available information of the whole disease course and tried predominantly to use the data about the onset of SJIA. The patients with flare were included in the present study only in cases in which the data of the disease onset were missing or insufficient. All the children were divided into 2 groups. The main group consisted of patients with active SJIA with definite MAS (n ¼ 18) and controls with active SJIA without MAS (n ¼ 40). If the patient had signs of MAS since admission or developed definite MAS later during this flare, he or she was referred to the main group. Only the children who did not have distinctive MAS during the flare episode and did not have MAS in the past medical history were referred to the control group—SJIA without MAS. The diagnosis of MAS was based on the criteria provided by A. Ravelli [24,27,28]. Each patient recognized with definite MAS was consistent to at least Ravelli's criteria or the HLH-2004 criteria and was reassessed by at least 3 experienced physicians (1 attending physician of the patient and at least 2 from the department staff). If there was no disagreement among the physicians and the patient was consistent to at least 1 of the 2 abovementioned criteria, the patient was included in the main group. Eight patients with probable/doubtful MAS (beyond the 58 patients remained in the study) were excluded if they fell under the criteria or disagreement between
physicians was reached. The disagreement was defined if the opinion of at least one of the physicians differed from the others. We collected the data about all the clinical and laboratory SJIArelated and MAS-related signs at the moment of the patient's admission to our clinic. The clinical signs were fever (Z 381C), rash, active joints, hepatosplenomegaly, jaundice, lymphadenopathy, cardio-respiratory involvement (pleuritis, respiratory failure, myocarditis, and pericarditis), central nervous system dysfunction (irritability, confusion, coma, seizures, and brain MRI patterns), hemorrhage syndrome, and bleeding. The number of active joints was recorded at the moment of admission to our clinic. Oligoarticular course means the patient has less than 5 active joints, if the number of active joints was more or equal to 5, the polyarticular course was implied. Among the laboratory signs, we checked hemoglobin (Hb), white blood cells (WBC), and platelets (PLT) count; erythrocyte sedimentation rate (ESR); C-reactive protein (CRP), ferritin, total serum protein, and albumin levels; triglycerides, prothrombin, fibrinogen, sodium and alanine aminotransferase, aspartate aminotransferase, γ-glutamintranspeptidase, and lactate dehydrogenase (LDH) activities; hyperbilirubinemia; and proteinuria. All blood samples were collected after fasting. We also evaluated the demographic data. The protocol of this study was approved by the local Ethic Committee of our University. Statistics The descriptive statistics were reported in terms of medians and interquartile ranges (IQRs) for continuous variables and in terms of absolute frequencies and percentages for categorical variables. We used the Mann–Whitney U-test for the comparison of quantitative variables in 2 groups and the chi-square test for the comparison of qualitative data or the Fisher's exact test in case of expected frequencies o5. The ability of each variable to discriminate the MAS episode from systemic arthritis flare was evaluated with sensitivity and specificity analysis, with area under receiver operating characteristic curve (AUC-ROC) analysis with 95% confidence interval (CI), and by calculating odds ratio (OR) for the detection the best cutoffs of continuous variables. The higher values of OR of variables interfere better discriminatory ability. For laboratory tests, we used AUC-ROC analysis with 95% CI. For each categorical variable the analysis of sensitivity and specificity was performed. We avoided using the known “standard” threshold (i.e., threshold reported in literature before or judged as clinically meaningful), because usually these tests better discriminate the advanced stage of MAS. We used the “best” threshold obtained through the ROC curve analysis of our data because they provide the most appropriate means between sensitivity and specificity. After selection of the most significant criteria, we tried to create the diagnostic rule based on combination of these criteria. The clinical and laboratory criteria and their combination were evaluated either separately or together. For calculation of number of possible combination, we used the “n of k” procedure. The software Statistica (release 6.0, StatSoft Corporation, Tulsa, OK), Biostat, and MedCalc were used for the data analyses. P o 0.05 was considered to indicate a significant difference.
Results The main demographic data on clinical and laboratory abnormalities in both groups are in Tables 1 and 2. The age of MAS onset was 6.7 (2.3–10.2) years, the interval between SJIA onset and MAS onset was 11.3 (0.8–34.8) months, and MAS duration was 68.0 (48.0–93.0) days, ranged: 11.0–336.0 days. The relapse course was experienced by 10/18 (55.6%) patients, and 2/18 patients died (11.1%). As the suspected children with MAS had typical changes,
M.M. Kostik et al. / Seminars in Arthritis and Rheumatism ] (2014) ]]]–]]]
3
Table 1 Demographic data and laboratory abnormalities in active SJIA children with and without MAS Signs
SJIA with MAS, Me (IQR)
SJIA without MAS, Me (IQR)
P value
Sex, females, n (%) Age of SJIA onset, years Active joints, n Hb, g/dl WBC 109/l PLT 109/l ESR, mm/h CRP, mg/l ALT, U/l AST, U/l GGTP, U/l ALP, U/l LDH, U/l Ferritin, μg/l Triglycerids, mmol/l Prothrombin, % Fibrinogen, g/l Total protein, g/l Albumin, g/l Na þ , mmol/l
11/18 3.8 3.0 8.8 6.1 108.5 10.0 125.0 113.0 244.0 53.0 347.0 1125.0 1798.5 1.995 70.5 1.4 59.2 24.2 137.0
21/40 4.35 8.0 10.6 18.2 532.0 43.0 53.7 34.9 36.7 28.9 339.5 512.0 143.0 1.21 95.0 5.6 71.7 41.0 139.0
0.54 0.64 0.004 0.001 0.0000001 0.0000001 0.0007 0.28 0.0003 0.000008 0.11 0.44 0.000006 0.00002 0.02 0.006 0.0008 0.00007 0.0000001 0.001
(61.1) (1.4; 6.6) (2.0; 6.0) (8.0; 10.0) (3.5; 9.2) (87.0; 200.0) (4.0; 38.0) (7.1; 159.0) (46.0; 383.5) (69.0; 489.0) (37.9; 302.3) (321.0; 797.5) (905.0; 1545.0) (1200.0; 10600.0) (1.53; 3.47) (53.0; 79.0) (0.8; 3.0) (49.0; 67.0) (23.0; 26.8) (135.0; 138.0)
(52.5) (2.6; 7.5) (4.0; 23.0) (9.5; 11.9) (12.4; 22.9) (408.0; 664.0) (34.0; 55.0) (18.0; 99.6) (25.9; 54.9) (31.6; 49.9) (20.9; 86.0) (258.5; 445.5) (420.5; 619.5) (117.0; 400.0) (0.96; 1.63) (86.0; 97.0) (3.9; 5.7) (68.0; 78.0) (36.0; 43.3) (138.0; 141.0)
ALP—alkaline phosphatase, ALT—alanine aminotransferase, AST—aspartate aminotransferase, CRP—C-reactive protein, ESR—erythrocyte sedimentation rate, GGTT—γglutamintranspeptidase, Hb—hemoglobin, IQR—interquartile range, LDH—lactate dehydrogenase, Me—median, Na þ —sodium, PLT—platelets, WBC—white blood cells.
realized in lower Hb, WBC, PLT, ESR, prothrombin, fibrinogen, total protein, albumin, and lower sodium levels. The ALT, AST, LDH, triglycerides, and ferritin levels were higher than in children with SJIA without the evidence of MAS. It is interesting that CRP was also higher in the MAS group, but the data were insignificant. Among the clinical signs, the patients with MAS had oligoarticular disease course, few number of active joints, active joints, spleen and liver enlargement, respiratory failure, and hemorrhages. There were no significant differences in such clinical symptoms as fever, rash, lymphadenopathy, and cardiac involvement. It is interesting that proteinuria during the SJIA flare was only in the MAS group. Proteinuria was only transient during the MAS episode and not associated with the consequent development of renal amyloidosis. The observation period after the proteinuria resolution was 6.5 (3.75–9.25) years, and no cases of amyloidosis have been observed up to now. The presence of articular involvement, particularly wrist, MCP, and PIP joints showed the protective effect of MAS development. The involvement of the cervical spine and the TMJ joints showed us borderline significant level too. The frequency of
hip involvement was nearly 3 times lower in the MAS group, but the data are insignificant. Some clinical signs, such as splenomegaly, lymphadenopathy, and hepatomegaly showed high sensitivity with poor specificity levels. Hemorrhages, respiratory failure, CNS dysfunction, and renal involvement revealed vice versa a high level of specificity with low sensitivity. Several clinical signs, wrist arthritis, MCP, and PIP joints, had the protective effect against MAS development. The best cutoffs for laboratory parameters related to MAS are in Table 3. The highest OR with high levels of sensitivity and specificity for the ability to discriminate MAS belonged to such variables as platelets count r 211 109/l, LDH 4 882 U/l (4 2 upper limits of norm), total protein r63 g/l, albumin r 2.9 g/dl, AST 4 59.7 U/l, fibrinogen r 1.8 g/l, and ferritin 4 400 μg/l. It is interesting that in our cohort we have revealed a lower ferritin level as cutoff than other researchers, possibly because we checked early laboratory changes for detecting the group of active SJIA who had early MAS or had an increased risk of MAS development and required a thorough future monitoring of the disease course.
Table 2 Presence of clinical signs in active SJIA children with and without MAS Signs
SJIA with MAS, n (%)
SJIA without MAS, n (%)
Sensitivity
Specificity
OR (95% CI)
P value
Fever Rash Oligoarthritis Splenomegaly Lymphadenopathy Hepatomegaly Hemorrhages Respiratory failure Cardiac involvement CNS dysfunction Renal involvement Wrist involvement MCP involvement PIP involvement Cervical spine involvement TMJ involvement Hip involvement
18/18 16/17 12/17 18/18 13/18 16/18 5/18 10/18 6/18 5/18 6/18 8/18 1/18 1/18 3/18 1/18 2/18
36/38 29/38 12/40 14/40 22/40 25/40 0/40 4/40 13/40 1/39 0/40 30/39 14/39 17/39 16/39 11/39 12/40
1.0 0.94 0.71 1.0 0.72 0.89 0.28 0.56 0.33 0.28 0.33 0.44 0.06 0.06 0.17 0.06 0.14
0.05 0.24 0.7 0.65 0.45 0.38 1.0 0.9 0.68 0.98 1.0 0.23 0.64 0.56 0.59 0.72 0.64
2.5 5.0 5.6 67.6 2.1 4.8 33.0 11.3 1.0 15.0 42.1 0.2 0.1 0.1 0.3 0.1 0.3
1.0a 0.14a 0.005 0.000004 0.21 0.04 0.002a 0.0001a 0.95 0.009a 0.0001a 0.016 0.02a 0.004 0.07 0.08a 0.19a
(100.0) (94.1) (70.6) (100.0) (72.2) (88.9) (27.8) (55.6) (33.3) (27.8) (33.3) (44.4) (5.6) (5.6) (16.7) (5.6) (11.1)
(94.7) (73.6) (30.0) (35.0) (55.0) (62.5) (0.0) (10.0) (32.5) (2.6) (0.0) (76.9) (35.9) (43.6) (41.0) (28.2) (30.0)
CNS—central nervous system, MCP—metacarpophalangeal joints, PIP—proximal interphalangeal joints, TMJ—temporomandibular joints. a
The Fisher's exact test.
(0.1–55.6) (0.6–42.8) (1.6–19.4) (3.8–1205.9) (0.6–7.1) (1.0–23.9) (1.7–636.8) (2.8–45.2) (0.3–3.4) (1.6–140.5) (2.2–801.2) (0.1–0.8) (0.0–0.9) (0.0–0.6) (0.1–1.2) (0.0–1.3) (0.1–1.5)
M.M. Kostik et al. / Seminars in Arthritis and Rheumatism ] (2014) ]]]–]]]
4
Table 3 The best cutoffs points specific for early recognition of MAS in active SJIA Parameter
Sensitivity
Specificity
OR (95% CI)a
AUC (95% CI)b
P value
Hb r 90 g/l WBC r 9.9 109/l PLT r 211 109/l ALT 4 72.9 U/l AST 4 59.7 U/l LDH 4 882 U/l GGTP 4 35 U/l ALP 4 736.2 Total protein r 63 g/l Albumin r 2.9 g/dl Prothrombin r 77% Fibrinogen r 1.8 g/l Ferritin 4 400 μg/l CRP 4 113 mg/l ESR r 10 mm/h TG 4 1.6 mmol/l Na þ r 137 mmol/l
72.2 83.3 88.9 64.7 82.4 75.0 83.3 27.3 64.7 100.0 71.4 64.3 100.0 52.9 61.1 71.4 66.7
80.0 90.0 100.0 85.0 92.1 100.0 60.0 100.0 97.5 92.5 92.3 100.0 76.0 89.5 92.5 79.2 81.1
10.4 35.0 534.6 10.4 54.4 158.3 7.5 30.1 71.5 375.0 30.0 46.6 87.0 9.6 19.4 6.1 8.0
0.77 0.92 0.98 0.81 0.88 0.91 0.68 0.58 0.84 0.98 0.81 0.88 0.92 0.59 0.78 0.73 0.79
0.0001 0.0000001 0.0000001c 0.0002 0.0000001 0.0000001c 0.047c 0.01c 0.0000001c 0.0000001 0.0008 0.001c 0.000005 0.001c 0.0001c 0.01 0.002c
(2.9–37.8) (7.4–165.6) (24.3–1747.8) (2.8–38.9) (9.8–302.9) (7.9–3169.1) (1.2–47.1) (1.4–638.4) (7.8–658.5) (18.4–7661.8) (2.9–313.5) (2.3–947.8) (4.5–1671.7) (2.3–39.1) (4.3–87.8) (1.4–26.0) (2.1–31.0)
(0.64–0.87) (0.81–0.97) (0.9–0.997) (0.68–0.9) (0.76–0.95) (0.79–0.976) (0.47–0.85) (0.43–0.72) (0.71–0.92) (0.9–0.997) (0.61–0.93) (0.7–0.97) (0.78–0.98) (0.45–0.72) (0.65–0.88) (0.56–0.86) (0.65–0.0.89)
ALP—alkaline phosphatase, ALT—alanine aminotransferase, AST—aspartate aminotransferase, CRP—C-reactive protein, ESR—erythrocyte sedimentation rate, GGTT—γglutamintranspeptidase, Hb—hemoglobin, IQR—interquartile range, LDH—lactate dehydrogenase, Me—median, Na þ —sodium, PLT—platelets, TG—triglycerides, WBC—white blood cells. a b c
OR—odds ratio, 95% CI—95% confidence interval. AUC—area under the curve, 95% CI—95% confidence interval. The Fisher's exact test.
Table 4 shows the results of diagnostic criteria selection, which can more precisely describe the early MAS development and discriminate it from SJIA. We evaluated the variables that provided strong discriminating properties and not duplicated with the others. We used variables with high OR and high levels of sensitivity and specificity. The clinical variables were splenomegaly, hemorrhages, CNS dysfunction, and respiratory failure; the laboratory variables included a decreased platelet count; a decreased white blood cells count; a decreased level of albumin, ferritin, and fibrinogen; an increased aspartate aminotransferase and lactate dehydrogenase activity; and the presence of proteinuria. We evaluated separately only clinical and laboratory variables and their combination. We calculated sensitivity, specificity, and OR for each combination. After the analysis, we revealed that only clinical variables and their combination (2 clinical variables) had the highest sensitivity level ¼ 0.68 and comparatively low OR ¼ 168.2 (8.9–3187.7). The laboratory variables were more precise in the discrimination of early MAS than clinical: any 3 or more laboratory criteria provided the highest specificity (1.0) and sensitivity (1.0) with the highest OR ¼ 2997 (57.2–156,962.8) in the whole model. Any attempt to combine clinical variables with laboratory variables does not change sensitivity, specificity, and the means of OR, so for early diagnosis of MAS, the laboratory criteria are more precise. The applying “n of k” procedure provided 56 possible combinations of a set of criteria. All possible Table 4 Selection of diagnostic criteria for early recognition of MAS in active SJIA Sensitivity
Specificity
Number of presenting laboratory criteriaa 1 And more 1.0 0.73 2 And more 1.0 0.95 3 And more 1.0 1.0 4 And more 0.94 1.0 5 And more 0.83 1.0 6 And more 0.5 1.0 7 And more 0.33 1.0 8 0.11 1.0
OR (95% CI)
96.7 599.4 2997.0 945.0 358.7 81.0 42.1 12.3
(5.5–1704.1) (27.4–13118.2) (57.2–156962.8) (36.7–24357.4) (17.5–7354.3) (4.3–1517.4) (2.2–801.2) (0.6–269.7)
a Laboratory criteria: PLT r 211 109/l, WBC r 9.9 109/l, AST 4 59.7 U/l, LDH 4 882 U/l, albumin r 2.9 g/dl, ferritin 4 400 μg/l, fibrinogen r 1.8 g/l, and proteinuria.
combinations were evaluated manually with the same sensitivity, specificity, and OR. Based on the data analysis, we created the preliminary diagnostic criteria (laboratory) for the early recognition of MAS in active SJIA.
Discussion MAS is a comparatively rare condition and is usually recognized by only experienced rheumatologists who work in big centers. The data of the International consensus survey of diagnostic criteria for macrophage activation syndrome in SJIA show the differences in the meaning of the MAS-related criteria between physicians from different continents. This problem can influence the results of multi-central international studies. The physicians from different continents pay attention to different symptoms: American physicians pay more attention to clinical signs and non-American and non-European physicians to hyperferritinemia. European physicians focus more on hypertriglyceridemia and less on falling ESR, prolongation of clothing time, and increased D-dimer [25]. The benefit of our study is based on the fact that all the patients included in the present study were from one pediatric department, so we used the same diagnostic methodology and rules that provided high homogeneity of our cohort. MAS can complicate different conditions in pediatric rheumatology, not only SJIA, but systemic lupus erythematosus, autoinflammatory disorders, and so on [29,30]. Currently there are only 2 sets of diagnostic criteria, the HLH criteria—more suitable for hereditary hematologic conditions and Ravelli's criteria, which were created for diagnosis of the MAS in SJIA. Both sets of criteria have limitations. For example, the HLH criteria are more susceptible in patients with severe cytopenia, which can be observed in SJIA children in the advanced stage, because the typical SJIA patients have leukocytosis and thrombocytosis due to hyperproduction of interleukin-6. Also, the HLH criteria are based on several points, which are not easy to perform for all SJIA patients, for example, assessment of NK-cell activity, CD-25 level, and evidence of hemophagocytosis [22]. The lack of opportunity to perform immunologic tests decreases the possibility to identify MAS, because it is necessary to have at least 5 criteria of the 8. Often SJIA patients showed a decreased number of NK cells, so this
M.M. Kostik et al. / Seminars in Arthritis and Rheumatism ] (2014) ]]]–]]] Table 5 Comparison of different sets of criteria for early recognition of MAS in our cohort of children with active SJIA Sensitivity Specificity OR (95% CI)a The HLH criteria (2004) [22]
0.22
Ravelli's criteria (2005) [24] 2 And more laboratory 0.89 2 And more clinical and/or 0.89 laboratory 3 And more clinical and/or 0.89 laboratory Our criteria 3 And more laboratory a
1.0
1.0
Acknowledgments We gratefully acknowledge the cooperation of the patients and their families involved in this study.
25.1 (1.3–496.3)
0.98 0.83
312.0 (26.4–3688.7) 37.7 (7.0–202.6)
0.95
152.0 (19.7–1175.1)
1.0
5
2997.0 (57.2–156,962.8)
OR—odds ratio, 95% CI—95% confidence interval.
criteria really does not work and should be excluded from the diagnostic set in the cases where it is necessary to discriminate MAS from active SJIA [31,32]. Ravelli's criteria are more sensitive and specific, namely, for identification of MAS in active SJIA patients, but one of the most known criteria ferritin has not been evaluated and so was not included in the diagnostic set [24]. The benefit of Ravelli's criteria is in the using of simple clinical and laboratory signs and symptoms, which are easy to perform in a large number of centers. The statement that morphological confirmation is necessary only in doubtful cases also makes the MAS diagnosis easier than applying the HLH criteria. Currently, there are no criteria for early MAS identification in SJIA children. It is known that in the HLH criteria, only 2/8 were clinical, and in Ravelli's criteria, only 3/8 were clinical. In the International consensus survey of diagnostic criteria for macrophage activation syndrome in SJIA there is only 1 clinical and 8 laboratory signs. Early suspicion of MAS is most commonly based on the detection of subtle laboratory changes, whereas clinical symptoms are often delayed. In our study, we show that clinical signs do not play such an important role in the early diagnosis of MAS compared with laboratory features, which are more sensitive and specific than clinical features. The comparison of different sets of diagnostic criteria is in Table 5. The established criteria of HLH have high specificity (1.0) with low sensitivity (0.2) and low diagnostic OR ¼ 25.1 (1.3– 496.3). The criteria, provided by Ravelli et al. [24] are more sensitive (0.89) and specific (0.83–0.95) and have higher diagnostic OR ¼ 37.7–312.0, depending on the number of variables. Our set of criteria has the highest sensitivity (1.0), specificity (1.0), and diagnostic OR ¼ 2997.0 (57.2–156,962.8), namely, for the identification of early MAS than previously described above. There are some limitations of our study related to its retrospective character, which lead to missing data, the absence of immunological evaluation and genetic testing, and a relatively small sample size of our cohort.
Conclusion In the biologic era, it is necessary to identify patients with commencing MAS to prevent exacerbation or development of serious adverse events because there are no sufficient data about the safety of some biologics in patients with SJIA and MAS. For this purpose, we tried to detect clinical and laboratory markers for the early identification of patients with active SJIA who have commenced MAS or have a high risk of MAS development. We created preliminary diagnostic (laboratory) guidelines able to early discriminate MAS in active SJIA. Our criteria are needed in prospective validation to evaluate their strength and future potential practical usefulness.
References [1] Prieur AM, Stephan JL. Macrophage activation syndrome in rheumatic diseases in children. Rev Rhum Ed Fr 1994;61:447–51. [2] Behrens EM, Beukelman T, Paessler M, Cron RQ. Occult macrophage activation syndrome in patients with systemic juvenile idiopathic arthritis. J Rheumatol 2007;34:1133–8. [3] Bleesing J, Prada A, Siegel DM, Villanueva J, Olson J, Ilowite NT, et al. The diagnostic significance of soluble CD163 and soluble interleukin-2 receptor alpha-chain in macrophage activation syndrome and untreated new-onset systemic juvenile idiopathic arthritis. Arthritis Rheum 2007;56:965–71. [4] Kelly A, Ramanan AV. Recognition and management of macrophage activation syndrome in juvenile arthritis. Curr Opin Rheumatol 2007;19:477–81. [5] Filipovich AH. Hemophagocytic lymphohistiocytosis (HLH) and related disorders. Hematology Am Soc Hematol Educ Program 2009:127–31. [6] Favara BE, Feller AC, Pauli M, Jaffe ES, Weiss LM, Arico M, et al. Contemporary classification of histiocytic disorders. The WHO Committee on Histiocytic/ Reticulum Cell Proliferations. Reclassification Working Group of the Histiocyte Society. Med Pediatr Oncol 1997;29:157–66. [7] Tristano AG, Casanova-Escalona L, Torres A, Rodriguez MA. Macrophage activation syndrome in a patient with systemic onset rheumatoid arthritis: rescue with intravenous immunoglobulin therapy. J Clin Rheumatol 2003;9: 253–8. [8] Miettunen PM, Narendran A, Jayanthan A, Behrens EM, Cron RQ. Successful treatment of severe paediatric rheumatic disease-associated macrophage activation syndrome with interleukin-1 inhibition following conventional immunosuppressive therapy: case series with 12 patients. Rheumatology (Oxford) 2011;50:417–9. [9] Bruck N, Suttorp M, Kabus M, Heubner G, Gahr M, Pessler F. Rapid and sustained remission of systemic juvenile idiopathic arthritis-associated macrophage activation syndrome through treatment with anakinra and corticosteroids. J Clin Rheumatol 2011;17:23–7. [10] Kelly A, Ramanan AV. A case of macrophage activation syndrome successfully treated with anakinra. Nat Clin Pract Rheumatol 2008;4:615–20. [11] Zhang K, Jordan MB, Marsh RA, Johnson JA, Kissell D, Meller J, et al. Hypomorphic mutations in PRF1, MUNC13-4, and STXBP2 are associated with adult-onset familial hemophagocytic lymphohistiocytosis. Blood 2011;118: 5794–8. [12] Sullivan KE, Delaat CA, Douglas SD, Filipovich AH. Defective natural killer cell function in patients with hemophagocytic lymphohistiocytosis and in first degree relatives. Pediatr Res 1998;44:465–8. [13] Feldmann J, Callebaut I, Raposo G, Certain S, Bacq D, Dumont C, et al. Munc13-4 is essential for cytolytic granules fusion and is mutated in a form of familial hemophagocytic lymphohistiocytosis (FHL3). Cell 2003;115:461–73. [14] Stepp SE, Dufourcq-Lagelouse R, Le Deist F, Bhawan S, Certain S, Mathew PA, et al. Perforin gene defects in familial hemophagocytic lymphohistiocytosis. Science 1999;286:1957–9. [15] zur Stadt U, Rohr J, Seifert W, Koch F, Grieve S, Pagel J, et al. Familial hemophagocytic lymphohistiocytosis type 5 (FHL-5) is caused by mutations in Munc18-2 and impaired binding to syntaxin 11. Am J Hum Genet 2009;85:482–92. [16] zur Stadt U, Schmidt S, Kasper B, Beutel K, Diler AS, Henter JI, et al. Linkage of familial hemophagocytic lymphohistiocytosis (FHL) type-4 to chromosome 6q24 and identification of mutations in syntaxin 11. Hum Mol Genet 2005;14:827–34. [17] Brown DE, McCoy MW, Pilonieta MC, Nix RN, Detweiler CS. Chronic murine typhoid fever is a natural model of secondary hemophagocytic lymphohistiocytosis. PLoS One 2010;5:e9441. [18] Yanagimachi M, Naruto T, Miyamae T, Hara T, Kikuchi M, Hara R, et al. Association of IRF5 polymorphisms with susceptibility to macrophage activation syndrome in patients with juvenile idiopathic arthritis. J Rheumatol 2011;38:769–74. [19] Behrens EM, Canna SW, Slade K, Rao S, Kreiger PA, Paessler M, et al. Repeated TLR9 stimulation results in macrophage activation syndrome-like disease in mice. J Clin Invest 2011;121:2264–77. [20] Fife MS, Gutierrez A, Ogilvie EM, Stock CJ, Samuel JM, Thomson W, et al. Novel IL10 gene family associations with systemic juvenile idiopathic arthritis. Arthritis Res Ther 2006;8:R148. [21] Moller JC, Paul D, Ganser G, Range U, Gahr M, Kelsch R, et al. IL10 promoter polymorphisms are associated with systemic onset juvenile idiopathic arthritis (SoJIA). Clin Exp Rheumatol 2010;28:912–8. [22] Henter JI, Horne A, Aricó M, Egeler RM, Filipovich AH, Imashuku S, et al. HLH-2004: Diagnostic and therapeutic guidelines for hemophagocytic lymphohistiocytosis. Pediatr Blood Cancer 2007;48:124–31. [23] Stephan J-L, Kone´-Paut I, Galanbrun C, Mouy R, Bader-Meunier B, Prieur A-M. Reactive haemophagocytic syndrome in children with inflammatory disorders: a retrospective study of 24 patients. Rheumatology 2001;40:1285–92.
6
M.M. Kostik et al. / Seminars in Arthritis and Rheumatism ] (2014) ]]]–]]]
[24] Ravelli A, Magni-Manzoni S, Pistorio A, Besana C, Foti T, Ruperto N, et al. Preliminary diagnostic guidelines for macrophage activation syndrome complicating systemic juvenile idiopathic arthritis. J Pediatr 2005;146:598–604. [25] Davì S1, Consolaro A, Guseinova D, Pistorio A, Ruperto N, Martini A, et al. An international consensus survey of diagnostic criteria for macrophage activation syndrome in systemic juvenile idiopathic arthritis. J Rheumatol 2011;38:764–8. [26] Petty RE, Southwood TR, Manners P, Baum J, Glass DN, Goldenberg J, et al. International League of Associations for Rheumatology classification of juvenile idiopathic arthritis: second revision, Edmonton 2001. J Rheumatol 2004;31:390–2. [27] Ravelli A. Macrophage activation syndrome. Curr Opin Rheumatol 2002;14: 548–52. [28] Ravelli A, Grom AA, Behrens EM, Cron RQ. Macrophage activation syndrome as part of systemic juvenile idiopathic arthritis: diagnosis, genetics, pathophysiology and treatment. Genes Immun 2012;13:289–98.
[29] Parodi A, Davi S, Pringe AB, Pistorio A, Ruperto N, Magni-Manzoni S, et al. Macrophage activation syndrome in juvenile systemic lupus erythematosus: a multinational multicenter study of thirty-eight patients. Arthritis Rheum 2009;60:3388–99. [30] Rigante D, Capoluongo E, Bertoni B, Ansuini V, Chiaretti A, Piastra M, et al. First report of macrophage activation syndrome in hyperimmunoglobulinemia D with periodic fever syndrome. Arthritis Rheum 2007;56:658–61. [31] de Jager W, Vastert SJ, Beekman JM, Wulffraat NM, Kuis W, Coffer PJ, et al. Defective phosphorylation of interleukin-18 receptor beta causes impaired natural killer cell function in systemic-onset juvenile idiopathic arthritis. Arthritis Rheum 2009;60:2782–93. [32] Zhou J, Tang X, Ding Y, An Y, Zhao X. Natural killer cell activity and frequency of killer cell immunoglobulin-like receptors in children with different forms of juvenile idiopathic arthritis. Pediatr Allergy Immunol 2013;24:691–6.
Contents lists available at ScienceDirect
Seminars in Arthritis and Rheumatism journal homepage: www.elsevier.com/locate/semarthrit
Identification of the best cutoff points and clinical signs specific for early recognition of macrophage activation syndrome in active systemic juvenile idiopathic arthritis Mikhail M. Kostik, MD, PhDn, Margarita F. Dubko, MD, PhD, Vera V. Masalova, MD, Ludmila S. Snegireva, MD, Tatyana L. Kornishina, MD, Irina A. Chikova, MD, Tatyana S. Likhacheva, MD, Eugenia A. Isupova, MD, Natalia I. Glebova, MD, Ekaterina M. Kuchinskaya, MD, Eugenia V. Balbotkina, MD, Natalia V. Buchinskaya, MD, Olga V. Kalashnikova, MD, PhD, Vyacheslav G. Chasnyk, MD, PhD Hospital Pediatry, Saint-Petersburg State Pediatric Medical University of Ministry of Healthcare of the Russia, 194100, Litovskaya 2, Saint-Petersburg, Russia
a r t i c l e in fo
Keywords: Systemic juvenile idiopathic arthritis Macrophage activation syndrome (MAS)
a b s t r a c t Objectives: The purpose of our study was to detect early clinical and laboratory signs that help to discriminate macrophage activation syndrome (MAS) from active systemic juvenile idiopathic arthritis (SJIA) without MAS. Methods: Our retrospective study was based on reviewing the medical charts of the children admitted to the rheumatology department with active SJIA and definite MAS (n ¼ 18) and without MAS (n ¼ 40). We evaluated the data related to SJIA and MAS at the moment of the patient's admission. If the patient had signs of MAS since admission or developed definite MAS later during this flare, he was referred to the main group. The children who did not have MAS during the flare episode and did not have MAS in the past medical history were in the control group. We calculated the cutoff points for MAS parameters, performed the analysis of sensitivity and specificity, identified the predictors, and provided the preliminary diagnostic rule through “the-number-of-criteria-present” approach. Results: The clinical signs were relevant to MAS in SJIA: oligoarticular disease course (OR ¼ 5.6), splenomegaly (OR ¼ 67.6), hemorrhages (OR ¼ 33.0), and respiratory failure (OR ¼ 11.3). The involvement of wrist (OR ¼ 0.2), MCP (OR ¼ 0.1), and PIP joints (OR ¼ 0.1) was protective against MAS development. The best cutoffs for laboratory parameters were PLT r 211 109/l, WBC r 9.9 109/l, AST 4 59.7 U/l, LDH 4 882 U/l, albumin r 2.9 g/dl, ferritin 4 400 μg/l, fibrinogen r 1.8 g/l, and proteinuria. The laboratory variables were more precise in the discrimination of early MAS than clinical: any 3 or more laboratory criteria provided the highest specificity (1.0) and sensitivity (1.0) and OR ¼ 2997. Conclusions: We detected clinical and laboratory markers and created preliminary diagnostic (laboratory) guidelines for early discrimination of MAS in active SJIA. & 2014 Elsevier Inc. All rights reserved.
Abbreviations: ALP, alkaline phosphatase; ALT, alanine aminotransferase; anti-IL1, anti-interleukin -1; anti-IL6, anti-interleukin-6; AST, aspartate aminotransferase; AUC, area under the curve; 95% CI, 95% confidence interval; CNS, central nervous system; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; GGTT, γ-glutamintranspeptidase; Hb, hemoglobin; HLH, hemophagocytic lymphohistiocytosis; ILAR, International League of Associations for Rheumatology; IQR, interquartile range; LDH, lactate dehydrogenase; MAS, macrophage activation syndrome; MCP, metacarpophalangeal joints; Me, median; Naþ , sodium; OR, odds ratio; PIP, proximal interphalangeous joints; PLT, platelets; SJIA, systemic juvenile idiopathic arthritis; TMJ, temporomandibular joints; TG, triglycerides; WBC, white blood cells. n Corresponding author. E-mail addresses: [email protected], [email protected] (M.M. Kostik). http://dx.doi.org/10.1016/j.semarthrit.2014.09.004 0049-0172/& 2014 Elsevier Inc. All rights reserved.
Introduction Macrophage activation syndrome (MAS) is a severe lifethreatening hematological condition, mostly complicated by systemic juvenile idiopathic arthritis (SJIA) [1]. The diagnosis of MAS can be difficult, especially in subclinical forms and in the early stage of MAS [2,3]. The early detection of MAS can lead to appropriate therapeutic interventions and change the outcomes. There are no strict criteria with enough sensitivity and specificity that are good for early MAS detection in SJIA [4]. MAS belongs to the family of histiocytic disorders, especially close to
2
M.M. Kostik et al. / Seminars in Arthritis and Rheumatism ] (2014) ]]]–]]]
hemophagocytic lymphohistiocytosis (HLH), but there are some differences in the pathogenesis between HLH and MAS that lead to different therapeutic approaches [5–10]. HLH is a typical genetic disorder related to T-cell and macrophage uncontrolled expansion linked to NK-cell and cytotoxic T-cell function considered to be the result of homozygous mutations in cytolytic pathway genes [11–16]. The pathogenesis of MAS is more complex and heterogenous. Decreased NK-cell activity, repeated stimulation, and hyperactivation of Toll-like receptor (TLR-9) and cytokine disbalance are other known pathogenic mechanisms underlying the MAS pathogenesis [17–21]. Despite the fact that MAS bears a close resemblance to hemophagocytic lymphohistiocytosis (HLH), the often applied HLH criteria created for genetic disorders are not applicable to early recognition of MAS in SJIA patients [22]. The prominent inflammation with an increased number of WBC and platelets and hyperfibrinogenemia typical for SJIA required a higher threshold of these parameters for MAS detection than in HLH. Fever, one of the main symptoms in SJIA and MAS/HLH, is very frequent in both conditions, so the sensitivity of fever is low. The main difference is the pattern of fever: non-remitting fever is typical for MAS/HLH and spiking remitting fever is more relevant for SJIA [23]. Due to the high intensity of thresholds of the HLH criteria they can only determinate the advanced stage of MAS, which leads to the delay in diagnosis and late start of specific treatment, which is associated with poor outcomes. Preliminary criteria of MAS in SJIA have been proposed, but these criteria had some limitations due to the lack of several laboratory measurements [24]. Several studies showed that clinical signs usually appear later and lead to the delay of MAS diagnosis, so the laboratory markers could be more sensitive and discriminative in early MAS diagnosis [24,25]. The purpose of our study was to detect clinical and laboratory signs that help to discriminate patients with an increased risk of MAS development or early MAS in active SJIA patients.
Methods Study design and patient selection Our retrospective study was based on reviewing the medical charts of the children who were admitted to our rheumatology department in 2005–2013 with the onset or flare of SJIA. The diagnosis of SJIA was established on ILAR definitions [26]. We collected all available information of the whole disease course and tried predominantly to use the data about the onset of SJIA. The patients with flare were included in the present study only in cases in which the data of the disease onset were missing or insufficient. All the children were divided into 2 groups. The main group consisted of patients with active SJIA with definite MAS (n ¼ 18) and controls with active SJIA without MAS (n ¼ 40). If the patient had signs of MAS since admission or developed definite MAS later during this flare, he or she was referred to the main group. Only the children who did not have distinctive MAS during the flare episode and did not have MAS in the past medical history were referred to the control group—SJIA without MAS. The diagnosis of MAS was based on the criteria provided by A. Ravelli [24,27,28]. Each patient recognized with definite MAS was consistent to at least Ravelli's criteria or the HLH-2004 criteria and was reassessed by at least 3 experienced physicians (1 attending physician of the patient and at least 2 from the department staff). If there was no disagreement among the physicians and the patient was consistent to at least 1 of the 2 abovementioned criteria, the patient was included in the main group. Eight patients with probable/doubtful MAS (beyond the 58 patients remained in the study) were excluded if they fell under the criteria or disagreement between
physicians was reached. The disagreement was defined if the opinion of at least one of the physicians differed from the others. We collected the data about all the clinical and laboratory SJIArelated and MAS-related signs at the moment of the patient's admission to our clinic. The clinical signs were fever (Z 381C), rash, active joints, hepatosplenomegaly, jaundice, lymphadenopathy, cardio-respiratory involvement (pleuritis, respiratory failure, myocarditis, and pericarditis), central nervous system dysfunction (irritability, confusion, coma, seizures, and brain MRI patterns), hemorrhage syndrome, and bleeding. The number of active joints was recorded at the moment of admission to our clinic. Oligoarticular course means the patient has less than 5 active joints, if the number of active joints was more or equal to 5, the polyarticular course was implied. Among the laboratory signs, we checked hemoglobin (Hb), white blood cells (WBC), and platelets (PLT) count; erythrocyte sedimentation rate (ESR); C-reactive protein (CRP), ferritin, total serum protein, and albumin levels; triglycerides, prothrombin, fibrinogen, sodium and alanine aminotransferase, aspartate aminotransferase, γ-glutamintranspeptidase, and lactate dehydrogenase (LDH) activities; hyperbilirubinemia; and proteinuria. All blood samples were collected after fasting. We also evaluated the demographic data. The protocol of this study was approved by the local Ethic Committee of our University. Statistics The descriptive statistics were reported in terms of medians and interquartile ranges (IQRs) for continuous variables and in terms of absolute frequencies and percentages for categorical variables. We used the Mann–Whitney U-test for the comparison of quantitative variables in 2 groups and the chi-square test for the comparison of qualitative data or the Fisher's exact test in case of expected frequencies o5. The ability of each variable to discriminate the MAS episode from systemic arthritis flare was evaluated with sensitivity and specificity analysis, with area under receiver operating characteristic curve (AUC-ROC) analysis with 95% confidence interval (CI), and by calculating odds ratio (OR) for the detection the best cutoffs of continuous variables. The higher values of OR of variables interfere better discriminatory ability. For laboratory tests, we used AUC-ROC analysis with 95% CI. For each categorical variable the analysis of sensitivity and specificity was performed. We avoided using the known “standard” threshold (i.e., threshold reported in literature before or judged as clinically meaningful), because usually these tests better discriminate the advanced stage of MAS. We used the “best” threshold obtained through the ROC curve analysis of our data because they provide the most appropriate means between sensitivity and specificity. After selection of the most significant criteria, we tried to create the diagnostic rule based on combination of these criteria. The clinical and laboratory criteria and their combination were evaluated either separately or together. For calculation of number of possible combination, we used the “n of k” procedure. The software Statistica (release 6.0, StatSoft Corporation, Tulsa, OK), Biostat, and MedCalc were used for the data analyses. P o 0.05 was considered to indicate a significant difference.
Results The main demographic data on clinical and laboratory abnormalities in both groups are in Tables 1 and 2. The age of MAS onset was 6.7 (2.3–10.2) years, the interval between SJIA onset and MAS onset was 11.3 (0.8–34.8) months, and MAS duration was 68.0 (48.0–93.0) days, ranged: 11.0–336.0 days. The relapse course was experienced by 10/18 (55.6%) patients, and 2/18 patients died (11.1%). As the suspected children with MAS had typical changes,
M.M. Kostik et al. / Seminars in Arthritis and Rheumatism ] (2014) ]]]–]]]
3
Table 1 Demographic data and laboratory abnormalities in active SJIA children with and without MAS Signs
SJIA with MAS, Me (IQR)
SJIA without MAS, Me (IQR)
P value
Sex, females, n (%) Age of SJIA onset, years Active joints, n Hb, g/dl WBC 109/l PLT 109/l ESR, mm/h CRP, mg/l ALT, U/l AST, U/l GGTP, U/l ALP, U/l LDH, U/l Ferritin, μg/l Triglycerids, mmol/l Prothrombin, % Fibrinogen, g/l Total protein, g/l Albumin, g/l Na þ , mmol/l
11/18 3.8 3.0 8.8 6.1 108.5 10.0 125.0 113.0 244.0 53.0 347.0 1125.0 1798.5 1.995 70.5 1.4 59.2 24.2 137.0
21/40 4.35 8.0 10.6 18.2 532.0 43.0 53.7 34.9 36.7 28.9 339.5 512.0 143.0 1.21 95.0 5.6 71.7 41.0 139.0
0.54 0.64 0.004 0.001 0.0000001 0.0000001 0.0007 0.28 0.0003 0.000008 0.11 0.44 0.000006 0.00002 0.02 0.006 0.0008 0.00007 0.0000001 0.001
(61.1) (1.4; 6.6) (2.0; 6.0) (8.0; 10.0) (3.5; 9.2) (87.0; 200.0) (4.0; 38.0) (7.1; 159.0) (46.0; 383.5) (69.0; 489.0) (37.9; 302.3) (321.0; 797.5) (905.0; 1545.0) (1200.0; 10600.0) (1.53; 3.47) (53.0; 79.0) (0.8; 3.0) (49.0; 67.0) (23.0; 26.8) (135.0; 138.0)
(52.5) (2.6; 7.5) (4.0; 23.0) (9.5; 11.9) (12.4; 22.9) (408.0; 664.0) (34.0; 55.0) (18.0; 99.6) (25.9; 54.9) (31.6; 49.9) (20.9; 86.0) (258.5; 445.5) (420.5; 619.5) (117.0; 400.0) (0.96; 1.63) (86.0; 97.0) (3.9; 5.7) (68.0; 78.0) (36.0; 43.3) (138.0; 141.0)
ALP—alkaline phosphatase, ALT—alanine aminotransferase, AST—aspartate aminotransferase, CRP—C-reactive protein, ESR—erythrocyte sedimentation rate, GGTT—γglutamintranspeptidase, Hb—hemoglobin, IQR—interquartile range, LDH—lactate dehydrogenase, Me—median, Na þ —sodium, PLT—platelets, WBC—white blood cells.
realized in lower Hb, WBC, PLT, ESR, prothrombin, fibrinogen, total protein, albumin, and lower sodium levels. The ALT, AST, LDH, triglycerides, and ferritin levels were higher than in children with SJIA without the evidence of MAS. It is interesting that CRP was also higher in the MAS group, but the data were insignificant. Among the clinical signs, the patients with MAS had oligoarticular disease course, few number of active joints, active joints, spleen and liver enlargement, respiratory failure, and hemorrhages. There were no significant differences in such clinical symptoms as fever, rash, lymphadenopathy, and cardiac involvement. It is interesting that proteinuria during the SJIA flare was only in the MAS group. Proteinuria was only transient during the MAS episode and not associated with the consequent development of renal amyloidosis. The observation period after the proteinuria resolution was 6.5 (3.75–9.25) years, and no cases of amyloidosis have been observed up to now. The presence of articular involvement, particularly wrist, MCP, and PIP joints showed the protective effect of MAS development. The involvement of the cervical spine and the TMJ joints showed us borderline significant level too. The frequency of
hip involvement was nearly 3 times lower in the MAS group, but the data are insignificant. Some clinical signs, such as splenomegaly, lymphadenopathy, and hepatomegaly showed high sensitivity with poor specificity levels. Hemorrhages, respiratory failure, CNS dysfunction, and renal involvement revealed vice versa a high level of specificity with low sensitivity. Several clinical signs, wrist arthritis, MCP, and PIP joints, had the protective effect against MAS development. The best cutoffs for laboratory parameters related to MAS are in Table 3. The highest OR with high levels of sensitivity and specificity for the ability to discriminate MAS belonged to such variables as platelets count r 211 109/l, LDH 4 882 U/l (4 2 upper limits of norm), total protein r63 g/l, albumin r 2.9 g/dl, AST 4 59.7 U/l, fibrinogen r 1.8 g/l, and ferritin 4 400 μg/l. It is interesting that in our cohort we have revealed a lower ferritin level as cutoff than other researchers, possibly because we checked early laboratory changes for detecting the group of active SJIA who had early MAS or had an increased risk of MAS development and required a thorough future monitoring of the disease course.
Table 2 Presence of clinical signs in active SJIA children with and without MAS Signs
SJIA with MAS, n (%)
SJIA without MAS, n (%)
Sensitivity
Specificity
OR (95% CI)
P value
Fever Rash Oligoarthritis Splenomegaly Lymphadenopathy Hepatomegaly Hemorrhages Respiratory failure Cardiac involvement CNS dysfunction Renal involvement Wrist involvement MCP involvement PIP involvement Cervical spine involvement TMJ involvement Hip involvement
18/18 16/17 12/17 18/18 13/18 16/18 5/18 10/18 6/18 5/18 6/18 8/18 1/18 1/18 3/18 1/18 2/18
36/38 29/38 12/40 14/40 22/40 25/40 0/40 4/40 13/40 1/39 0/40 30/39 14/39 17/39 16/39 11/39 12/40
1.0 0.94 0.71 1.0 0.72 0.89 0.28 0.56 0.33 0.28 0.33 0.44 0.06 0.06 0.17 0.06 0.14
0.05 0.24 0.7 0.65 0.45 0.38 1.0 0.9 0.68 0.98 1.0 0.23 0.64 0.56 0.59 0.72 0.64
2.5 5.0 5.6 67.6 2.1 4.8 33.0 11.3 1.0 15.0 42.1 0.2 0.1 0.1 0.3 0.1 0.3
1.0a 0.14a 0.005 0.000004 0.21 0.04 0.002a 0.0001a 0.95 0.009a 0.0001a 0.016 0.02a 0.004 0.07 0.08a 0.19a
(100.0) (94.1) (70.6) (100.0) (72.2) (88.9) (27.8) (55.6) (33.3) (27.8) (33.3) (44.4) (5.6) (5.6) (16.7) (5.6) (11.1)
(94.7) (73.6) (30.0) (35.0) (55.0) (62.5) (0.0) (10.0) (32.5) (2.6) (0.0) (76.9) (35.9) (43.6) (41.0) (28.2) (30.0)
CNS—central nervous system, MCP—metacarpophalangeal joints, PIP—proximal interphalangeal joints, TMJ—temporomandibular joints. a
The Fisher's exact test.
(0.1–55.6) (0.6–42.8) (1.6–19.4) (3.8–1205.9) (0.6–7.1) (1.0–23.9) (1.7–636.8) (2.8–45.2) (0.3–3.4) (1.6–140.5) (2.2–801.2) (0.1–0.8) (0.0–0.9) (0.0–0.6) (0.1–1.2) (0.0–1.3) (0.1–1.5)
M.M. Kostik et al. / Seminars in Arthritis and Rheumatism ] (2014) ]]]–]]]
4
Table 3 The best cutoffs points specific for early recognition of MAS in active SJIA Parameter
Sensitivity
Specificity
OR (95% CI)a
AUC (95% CI)b
P value
Hb r 90 g/l WBC r 9.9 109/l PLT r 211 109/l ALT 4 72.9 U/l AST 4 59.7 U/l LDH 4 882 U/l GGTP 4 35 U/l ALP 4 736.2 Total protein r 63 g/l Albumin r 2.9 g/dl Prothrombin r 77% Fibrinogen r 1.8 g/l Ferritin 4 400 μg/l CRP 4 113 mg/l ESR r 10 mm/h TG 4 1.6 mmol/l Na þ r 137 mmol/l
72.2 83.3 88.9 64.7 82.4 75.0 83.3 27.3 64.7 100.0 71.4 64.3 100.0 52.9 61.1 71.4 66.7
80.0 90.0 100.0 85.0 92.1 100.0 60.0 100.0 97.5 92.5 92.3 100.0 76.0 89.5 92.5 79.2 81.1
10.4 35.0 534.6 10.4 54.4 158.3 7.5 30.1 71.5 375.0 30.0 46.6 87.0 9.6 19.4 6.1 8.0
0.77 0.92 0.98 0.81 0.88 0.91 0.68 0.58 0.84 0.98 0.81 0.88 0.92 0.59 0.78 0.73 0.79
0.0001 0.0000001 0.0000001c 0.0002 0.0000001 0.0000001c 0.047c 0.01c 0.0000001c 0.0000001 0.0008 0.001c 0.000005 0.001c 0.0001c 0.01 0.002c
(2.9–37.8) (7.4–165.6) (24.3–1747.8) (2.8–38.9) (9.8–302.9) (7.9–3169.1) (1.2–47.1) (1.4–638.4) (7.8–658.5) (18.4–7661.8) (2.9–313.5) (2.3–947.8) (4.5–1671.7) (2.3–39.1) (4.3–87.8) (1.4–26.0) (2.1–31.0)
(0.64–0.87) (0.81–0.97) (0.9–0.997) (0.68–0.9) (0.76–0.95) (0.79–0.976) (0.47–0.85) (0.43–0.72) (0.71–0.92) (0.9–0.997) (0.61–0.93) (0.7–0.97) (0.78–0.98) (0.45–0.72) (0.65–0.88) (0.56–0.86) (0.65–0.0.89)
ALP—alkaline phosphatase, ALT—alanine aminotransferase, AST—aspartate aminotransferase, CRP—C-reactive protein, ESR—erythrocyte sedimentation rate, GGTT—γglutamintranspeptidase, Hb—hemoglobin, IQR—interquartile range, LDH—lactate dehydrogenase, Me—median, Na þ —sodium, PLT—platelets, TG—triglycerides, WBC—white blood cells. a b c
OR—odds ratio, 95% CI—95% confidence interval. AUC—area under the curve, 95% CI—95% confidence interval. The Fisher's exact test.
Table 4 shows the results of diagnostic criteria selection, which can more precisely describe the early MAS development and discriminate it from SJIA. We evaluated the variables that provided strong discriminating properties and not duplicated with the others. We used variables with high OR and high levels of sensitivity and specificity. The clinical variables were splenomegaly, hemorrhages, CNS dysfunction, and respiratory failure; the laboratory variables included a decreased platelet count; a decreased white blood cells count; a decreased level of albumin, ferritin, and fibrinogen; an increased aspartate aminotransferase and lactate dehydrogenase activity; and the presence of proteinuria. We evaluated separately only clinical and laboratory variables and their combination. We calculated sensitivity, specificity, and OR for each combination. After the analysis, we revealed that only clinical variables and their combination (2 clinical variables) had the highest sensitivity level ¼ 0.68 and comparatively low OR ¼ 168.2 (8.9–3187.7). The laboratory variables were more precise in the discrimination of early MAS than clinical: any 3 or more laboratory criteria provided the highest specificity (1.0) and sensitivity (1.0) with the highest OR ¼ 2997 (57.2–156,962.8) in the whole model. Any attempt to combine clinical variables with laboratory variables does not change sensitivity, specificity, and the means of OR, so for early diagnosis of MAS, the laboratory criteria are more precise. The applying “n of k” procedure provided 56 possible combinations of a set of criteria. All possible Table 4 Selection of diagnostic criteria for early recognition of MAS in active SJIA Sensitivity
Specificity
Number of presenting laboratory criteriaa 1 And more 1.0 0.73 2 And more 1.0 0.95 3 And more 1.0 1.0 4 And more 0.94 1.0 5 And more 0.83 1.0 6 And more 0.5 1.0 7 And more 0.33 1.0 8 0.11 1.0
OR (95% CI)
96.7 599.4 2997.0 945.0 358.7 81.0 42.1 12.3
(5.5–1704.1) (27.4–13118.2) (57.2–156962.8) (36.7–24357.4) (17.5–7354.3) (4.3–1517.4) (2.2–801.2) (0.6–269.7)
a Laboratory criteria: PLT r 211 109/l, WBC r 9.9 109/l, AST 4 59.7 U/l, LDH 4 882 U/l, albumin r 2.9 g/dl, ferritin 4 400 μg/l, fibrinogen r 1.8 g/l, and proteinuria.
combinations were evaluated manually with the same sensitivity, specificity, and OR. Based on the data analysis, we created the preliminary diagnostic criteria (laboratory) for the early recognition of MAS in active SJIA.
Discussion MAS is a comparatively rare condition and is usually recognized by only experienced rheumatologists who work in big centers. The data of the International consensus survey of diagnostic criteria for macrophage activation syndrome in SJIA show the differences in the meaning of the MAS-related criteria between physicians from different continents. This problem can influence the results of multi-central international studies. The physicians from different continents pay attention to different symptoms: American physicians pay more attention to clinical signs and non-American and non-European physicians to hyperferritinemia. European physicians focus more on hypertriglyceridemia and less on falling ESR, prolongation of clothing time, and increased D-dimer [25]. The benefit of our study is based on the fact that all the patients included in the present study were from one pediatric department, so we used the same diagnostic methodology and rules that provided high homogeneity of our cohort. MAS can complicate different conditions in pediatric rheumatology, not only SJIA, but systemic lupus erythematosus, autoinflammatory disorders, and so on [29,30]. Currently there are only 2 sets of diagnostic criteria, the HLH criteria—more suitable for hereditary hematologic conditions and Ravelli's criteria, which were created for diagnosis of the MAS in SJIA. Both sets of criteria have limitations. For example, the HLH criteria are more susceptible in patients with severe cytopenia, which can be observed in SJIA children in the advanced stage, because the typical SJIA patients have leukocytosis and thrombocytosis due to hyperproduction of interleukin-6. Also, the HLH criteria are based on several points, which are not easy to perform for all SJIA patients, for example, assessment of NK-cell activity, CD-25 level, and evidence of hemophagocytosis [22]. The lack of opportunity to perform immunologic tests decreases the possibility to identify MAS, because it is necessary to have at least 5 criteria of the 8. Often SJIA patients showed a decreased number of NK cells, so this
M.M. Kostik et al. / Seminars in Arthritis and Rheumatism ] (2014) ]]]–]]] Table 5 Comparison of different sets of criteria for early recognition of MAS in our cohort of children with active SJIA Sensitivity Specificity OR (95% CI)a The HLH criteria (2004) [22]
0.22
Ravelli's criteria (2005) [24] 2 And more laboratory 0.89 2 And more clinical and/or 0.89 laboratory 3 And more clinical and/or 0.89 laboratory Our criteria 3 And more laboratory a
1.0
1.0
Acknowledgments We gratefully acknowledge the cooperation of the patients and their families involved in this study.
25.1 (1.3–496.3)
0.98 0.83
312.0 (26.4–3688.7) 37.7 (7.0–202.6)
0.95
152.0 (19.7–1175.1)
1.0
5
2997.0 (57.2–156,962.8)
OR—odds ratio, 95% CI—95% confidence interval.
criteria really does not work and should be excluded from the diagnostic set in the cases where it is necessary to discriminate MAS from active SJIA [31,32]. Ravelli's criteria are more sensitive and specific, namely, for identification of MAS in active SJIA patients, but one of the most known criteria ferritin has not been evaluated and so was not included in the diagnostic set [24]. The benefit of Ravelli's criteria is in the using of simple clinical and laboratory signs and symptoms, which are easy to perform in a large number of centers. The statement that morphological confirmation is necessary only in doubtful cases also makes the MAS diagnosis easier than applying the HLH criteria. Currently, there are no criteria for early MAS identification in SJIA children. It is known that in the HLH criteria, only 2/8 were clinical, and in Ravelli's criteria, only 3/8 were clinical. In the International consensus survey of diagnostic criteria for macrophage activation syndrome in SJIA there is only 1 clinical and 8 laboratory signs. Early suspicion of MAS is most commonly based on the detection of subtle laboratory changes, whereas clinical symptoms are often delayed. In our study, we show that clinical signs do not play such an important role in the early diagnosis of MAS compared with laboratory features, which are more sensitive and specific than clinical features. The comparison of different sets of diagnostic criteria is in Table 5. The established criteria of HLH have high specificity (1.0) with low sensitivity (0.2) and low diagnostic OR ¼ 25.1 (1.3– 496.3). The criteria, provided by Ravelli et al. [24] are more sensitive (0.89) and specific (0.83–0.95) and have higher diagnostic OR ¼ 37.7–312.0, depending on the number of variables. Our set of criteria has the highest sensitivity (1.0), specificity (1.0), and diagnostic OR ¼ 2997.0 (57.2–156,962.8), namely, for the identification of early MAS than previously described above. There are some limitations of our study related to its retrospective character, which lead to missing data, the absence of immunological evaluation and genetic testing, and a relatively small sample size of our cohort.
Conclusion In the biologic era, it is necessary to identify patients with commencing MAS to prevent exacerbation or development of serious adverse events because there are no sufficient data about the safety of some biologics in patients with SJIA and MAS. For this purpose, we tried to detect clinical and laboratory markers for the early identification of patients with active SJIA who have commenced MAS or have a high risk of MAS development. We created preliminary diagnostic (laboratory) guidelines able to early discriminate MAS in active SJIA. Our criteria are needed in prospective validation to evaluate their strength and future potential practical usefulness.
References [1] Prieur AM, Stephan JL. Macrophage activation syndrome in rheumatic diseases in children. Rev Rhum Ed Fr 1994;61:447–51. [2] Behrens EM, Beukelman T, Paessler M, Cron RQ. Occult macrophage activation syndrome in patients with systemic juvenile idiopathic arthritis. J Rheumatol 2007;34:1133–8. [3] Bleesing J, Prada A, Siegel DM, Villanueva J, Olson J, Ilowite NT, et al. The diagnostic significance of soluble CD163 and soluble interleukin-2 receptor alpha-chain in macrophage activation syndrome and untreated new-onset systemic juvenile idiopathic arthritis. Arthritis Rheum 2007;56:965–71. [4] Kelly A, Ramanan AV. Recognition and management of macrophage activation syndrome in juvenile arthritis. Curr Opin Rheumatol 2007;19:477–81. [5] Filipovich AH. Hemophagocytic lymphohistiocytosis (HLH) and related disorders. Hematology Am Soc Hematol Educ Program 2009:127–31. [6] Favara BE, Feller AC, Pauli M, Jaffe ES, Weiss LM, Arico M, et al. Contemporary classification of histiocytic disorders. The WHO Committee on Histiocytic/ Reticulum Cell Proliferations. Reclassification Working Group of the Histiocyte Society. Med Pediatr Oncol 1997;29:157–66. [7] Tristano AG, Casanova-Escalona L, Torres A, Rodriguez MA. Macrophage activation syndrome in a patient with systemic onset rheumatoid arthritis: rescue with intravenous immunoglobulin therapy. J Clin Rheumatol 2003;9: 253–8. [8] Miettunen PM, Narendran A, Jayanthan A, Behrens EM, Cron RQ. Successful treatment of severe paediatric rheumatic disease-associated macrophage activation syndrome with interleukin-1 inhibition following conventional immunosuppressive therapy: case series with 12 patients. Rheumatology (Oxford) 2011;50:417–9. [9] Bruck N, Suttorp M, Kabus M, Heubner G, Gahr M, Pessler F. Rapid and sustained remission of systemic juvenile idiopathic arthritis-associated macrophage activation syndrome through treatment with anakinra and corticosteroids. J Clin Rheumatol 2011;17:23–7. [10] Kelly A, Ramanan AV. A case of macrophage activation syndrome successfully treated with anakinra. Nat Clin Pract Rheumatol 2008;4:615–20. [11] Zhang K, Jordan MB, Marsh RA, Johnson JA, Kissell D, Meller J, et al. Hypomorphic mutations in PRF1, MUNC13-4, and STXBP2 are associated with adult-onset familial hemophagocytic lymphohistiocytosis. Blood 2011;118: 5794–8. [12] Sullivan KE, Delaat CA, Douglas SD, Filipovich AH. Defective natural killer cell function in patients with hemophagocytic lymphohistiocytosis and in first degree relatives. Pediatr Res 1998;44:465–8. [13] Feldmann J, Callebaut I, Raposo G, Certain S, Bacq D, Dumont C, et al. Munc13-4 is essential for cytolytic granules fusion and is mutated in a form of familial hemophagocytic lymphohistiocytosis (FHL3). Cell 2003;115:461–73. [14] Stepp SE, Dufourcq-Lagelouse R, Le Deist F, Bhawan S, Certain S, Mathew PA, et al. Perforin gene defects in familial hemophagocytic lymphohistiocytosis. Science 1999;286:1957–9. [15] zur Stadt U, Rohr J, Seifert W, Koch F, Grieve S, Pagel J, et al. Familial hemophagocytic lymphohistiocytosis type 5 (FHL-5) is caused by mutations in Munc18-2 and impaired binding to syntaxin 11. Am J Hum Genet 2009;85:482–92. [16] zur Stadt U, Schmidt S, Kasper B, Beutel K, Diler AS, Henter JI, et al. Linkage of familial hemophagocytic lymphohistiocytosis (FHL) type-4 to chromosome 6q24 and identification of mutations in syntaxin 11. Hum Mol Genet 2005;14:827–34. [17] Brown DE, McCoy MW, Pilonieta MC, Nix RN, Detweiler CS. Chronic murine typhoid fever is a natural model of secondary hemophagocytic lymphohistiocytosis. PLoS One 2010;5:e9441. [18] Yanagimachi M, Naruto T, Miyamae T, Hara T, Kikuchi M, Hara R, et al. Association of IRF5 polymorphisms with susceptibility to macrophage activation syndrome in patients with juvenile idiopathic arthritis. J Rheumatol 2011;38:769–74. [19] Behrens EM, Canna SW, Slade K, Rao S, Kreiger PA, Paessler M, et al. Repeated TLR9 stimulation results in macrophage activation syndrome-like disease in mice. J Clin Invest 2011;121:2264–77. [20] Fife MS, Gutierrez A, Ogilvie EM, Stock CJ, Samuel JM, Thomson W, et al. Novel IL10 gene family associations with systemic juvenile idiopathic arthritis. Arthritis Res Ther 2006;8:R148. [21] Moller JC, Paul D, Ganser G, Range U, Gahr M, Kelsch R, et al. IL10 promoter polymorphisms are associated with systemic onset juvenile idiopathic arthritis (SoJIA). Clin Exp Rheumatol 2010;28:912–8. [22] Henter JI, Horne A, Aricó M, Egeler RM, Filipovich AH, Imashuku S, et al. HLH-2004: Diagnostic and therapeutic guidelines for hemophagocytic lymphohistiocytosis. Pediatr Blood Cancer 2007;48:124–31. [23] Stephan J-L, Kone´-Paut I, Galanbrun C, Mouy R, Bader-Meunier B, Prieur A-M. Reactive haemophagocytic syndrome in children with inflammatory disorders: a retrospective study of 24 patients. Rheumatology 2001;40:1285–92.
6
M.M. Kostik et al. / Seminars in Arthritis and Rheumatism ] (2014) ]]]–]]]
[24] Ravelli A, Magni-Manzoni S, Pistorio A, Besana C, Foti T, Ruperto N, et al. Preliminary diagnostic guidelines for macrophage activation syndrome complicating systemic juvenile idiopathic arthritis. J Pediatr 2005;146:598–604. [25] Davì S1, Consolaro A, Guseinova D, Pistorio A, Ruperto N, Martini A, et al. An international consensus survey of diagnostic criteria for macrophage activation syndrome in systemic juvenile idiopathic arthritis. J Rheumatol 2011;38:764–8. [26] Petty RE, Southwood TR, Manners P, Baum J, Glass DN, Goldenberg J, et al. International League of Associations for Rheumatology classification of juvenile idiopathic arthritis: second revision, Edmonton 2001. J Rheumatol 2004;31:390–2. [27] Ravelli A. Macrophage activation syndrome. Curr Opin Rheumatol 2002;14: 548–52. [28] Ravelli A, Grom AA, Behrens EM, Cron RQ. Macrophage activation syndrome as part of systemic juvenile idiopathic arthritis: diagnosis, genetics, pathophysiology and treatment. Genes Immun 2012;13:289–98.
[29] Parodi A, Davi S, Pringe AB, Pistorio A, Ruperto N, Magni-Manzoni S, et al. Macrophage activation syndrome in juvenile systemic lupus erythematosus: a multinational multicenter study of thirty-eight patients. Arthritis Rheum 2009;60:3388–99. [30] Rigante D, Capoluongo E, Bertoni B, Ansuini V, Chiaretti A, Piastra M, et al. First report of macrophage activation syndrome in hyperimmunoglobulinemia D with periodic fever syndrome. Arthritis Rheum 2007;56:658–61. [31] de Jager W, Vastert SJ, Beekman JM, Wulffraat NM, Kuis W, Coffer PJ, et al. Defective phosphorylation of interleukin-18 receptor beta causes impaired natural killer cell function in systemic-onset juvenile idiopathic arthritis. Arthritis Rheum 2009;60:2782–93. [32] Zhou J, Tang X, Ding Y, An Y, Zhao X. Natural killer cell activity and frequency of killer cell immunoglobulin-like receptors in children with different forms of juvenile idiopathic arthritis. Pediatr Allergy Immunol 2013;24:691–6.