DOI:
Ahead of print
Ultrasound versus physical examination in predicting disease flare in children with juvenile idiopathic arthritis: a systematic literature review and qualitative synthesis
Orazio De Lucia
1*, Teresa Giani
2,3*, Roberto Caporali
1,4,5, Rolando Cimaz
4,5,6* the authors share the first authorship
1Clinical Rheumatology Unit, Department of Rheumatology and Medical Sciences, ASST Centro Traumatologico Ortopedico G. Pini-CTO, Milano, 2AOU Meyer, Florence, Italy, 3Department of Medical Biotechnology, University of Siena, Siena, 4Department of Clinical Sciences and Community Health, University of Milano, 5REsearch Center for Adult and Pediatric Rheumatic Diseases, Milano, 6Pediatric Rheumatology Unit, Department of Rheumatology and Medical Sciences, ASST Centro Traumatologico Ortopedico G. Pini-CTO, Milano, Italy
Received 06.08.2021 Accepted 20.10.2021 Med Ultrason
2022:0 Online first, 1-6
Corresponding author: Orazio De Lucia, MD PhD
Department of Rheumatology and Medical Sciences, Clinical Rheumatology Unit ASST Centro Specialistico Ortopedico Traumatologico Gaetano Pini-CTO Piazza Cardinal Ferrari, 1 – 20122 Milano Phone: 0258296272
E-mail: [email protected]
Introduction
Ultrasound (US) has gained widespread usage, since it is safe, cheap, easily available, non-invasive, free of ionizing radiations and able to allow a real-time scanning evaluation at bedside. In the last few years technologi- cal advances have improved the ability of US to visual- ize more superficial and deeper areas and to detect blood
flow through small vessels, making it suitable for muscu- loskeletal applications and strongly increasing its use in rheumatology. US can closely image articular and peri- articular structures and has become part of daily clinical practice for many rheumatologists. However, in growing subjects physiologic variants can create subtle pitfalls [1,2], and training with use of advanced equipment are necessary to limit potential misinterpretations [3].
Despite the expanding availability of this tool and be- yond its technological limitations, a crucial question is its role in decision making, strictly linked to the expecta- tions related to the use in routine care. The assessment of active arthritis and procedural guidance appear to be the main applications in the field, but if the support of US as a procedural guidance during arthrocentesis and corticos- teroid injections is of proven utility, more controversial is its added value with respect to clinical examination in the detection of arthritis [4]. Even more intriguing for clini- Abstract
In this systematic review we analyzed the published articles related to the predictive value for flare of subclinical synovitis assessed by ultrasound (US) in juvenile idiopathic arthritis (JIA). Medline, Embase and Cochrane databases were searched from 1990 to 2020 by two authors, using PICO methodology. The study is built and reported according to PRISMA guidelines.
Searches identified four articles comprising a total of 187 JIA patients in clinical remission from at least 3 months. Two of the articles found US subclinical signs of synovitis to be predictive for flare, with a five times higher risk (with Power Doppler signal as an important feature), while in the other two baseline US abnormalities did not predict a clinical flare. The articles differed for protocols, definitions, and length of follow-up. US has an expanding role in pediatric rheumatology, with interest- ing applications especially during the follow-up, potentially identifying subclinical inflammatory signs predictive of flare.
However, the few studies available do not allow definite conclusions at this time.
Keywords: ultrasound; Juvenile Idiopathic Arthritis; flare; remission; subclinical synovitis
DOI: 10.11152/mu-3303
cians is its potential contribution in treatment manage- ment [5]. Several studies have shown a high sensitivity of US for minimal signs of inflammation even in patients in apparent clinical remission [6-10], but the meaning of subclinical synovitis in terms of risk of flare, which would be crucial for the definition of treatment strategies is still to be defined [11]. In fact, in a study performed on rheumatoid arthritis (RA) patients who underwent a step-up disease-modifying antirheumatic drug escala- tion guided by US results, there was no significant im- provement in outcome [12]. Other studies related to this topic have yielded conflicting results [13].
When considering juvenile idiopathic arthritis (JIA), a high percentage of paediatric patients show arthritis flare upon withdrawal of therapy and the availability of a tool able to guide decision-making related to stop- ping treatment strategies would be certainly needed [14].
However, in children the higher risk of US misinterpreta- tions mostly linked to the anatomical peculiarities of a growing skeleton make the real significance of suspected findings still less clear than in adults. There have been few studies that have explored the ability of US to predict clinical flares in JIA. Our aim was to review published articles related to the predictive value of subclinical synovitis for arthritis flare in children affected by JIA in clinical remission.
Material and methods
The systematic review was conducted according to the recommendations of PRISMA (Preferred Reporting Items for Systematic Review and Meta-Analysis) as con- veyed in Moher’s guidelines [15]. The search question was if subclinical synovitis assessed by US is able to pre- dict flare in children with JIA in clinical remission.
Medline, Embase and Cochrane databases were searched from 1990 to 2020. The search terms entered were: Juvenile Idiopathic Arthritis (JIA), Juvenile Rheu- matoid Arthritis (JRA), Juvenile Chronic Arthritis (JCA),
juvenile, child, children, adolescents, teenagers, youth, remission, clinical inactive disease as population, US as intervention, and subclinical synovitis, predictive value, flare, relapse as outcomes. All relevant index and natural language terms were tailored for all databases searched.
In addition, further relevant references were manually searched if needed. The inclusion and exclusion criteria are summarized in table I.
After removal of duplicate results, the first screen- ing process was done considering the title and abstract, and a further selection was done based on the full text.
The review selection was independently performed by two reviewers (ODL and TG). The discrepancies were solved by discussion with a third reviewer (RC) involved in case of no consensus achieved. The information about the numbers of articles generated by using search terms, the numbers of articles ruled out after the first screening and the reason for any excluded article were inserted in a PRISMA flow chart.
Quality assessment
The articles that met all inclusion criteria were evalu- ated in relation to methodological quality. The Quality Assessment Tools used for the selected studies were the Consolidated Standards of Reporting Trials (CONSORT Statement; http://www.consort-statement.org/) [16] and the Quality in Prognosis Studies tool. 5-point Oxford Quality Rating Scale [17].
Results
The search found 208 records. After duplicate remov- als 168 records were reviewed on title and abstract and, of those, 162 were excluded. Six records were assessed for eligibility and 2 were excluded after full text reading 1 for wrong outcome and the other for insufficient data (fig 1). The final result consisted therefore of 4 articles, which are summarized in Table II.
The 4 articles [18-21] were published in the last 7 years and include a total of 202 patients (88 in the larger Table I. Inclusion /exclusion criteria for screening titles and abstracts.
Inclusion criteria Exclusion criteria
• Population: Diagnosis of juvenile idiopathic arthritis based on ILAR criteria; Age <16 years; Clinical remission of JIA
• Study design: Randomised controlled trials (RCTs);
Systematic Reviews; Prospective studies;
Case series (> 10 patients)
• Index test: Ultrasound plus clinical assessment of any joint
• Reference: Clinical diagnosis of disease flare
• Outcome: Clinical flare within 12 months
• Abstract only, Letters to Editor, Case reports (≤ 10 patients);
• When multiple articles were based on the same study population, we included only the most complete (or recent) one.
one). Age at baseline was homogenous (medians 10, 10, 11, and 11 years). In 3/4 studies the most prevalent ILAR category was persistent oligoarticular, while in the study by Zhao et al [21] RF-negative polyarticular JIA account- ed for 38% of cases.
At the time of enrolment, number of patients off med- ications varied, being 57% in Magni Manzoni’s analysis [20], 34% in De Lucia’s, 25.7% in Miotto’s [19] and 13%
in Zhao’s [21]. Methotrexate was the most used drug in patients under treatment.
Two of the studies [18,19] showed an increased risk of flare in patients who were US-positive at baseline, while the other two [20,21] could not show such predic- tive value. Interestingly, in the first two studies the risk of flare was similar, being about five times higher for patients who showed a positive baseline US when com- Table II. Summary of the selected articles
Study De Lucia O [18] Miotto V [19] Magni Manzoni S
[20] Zhao Y [21]
Publication date and
study type
2018Prospective, case control, multicentric
2017Prospective, monocentric
2013Prospective, case control, monocentric
2018Prospective, monocentric Clinically inactive patients:
age (yrs) at baseline 10±4.3 clinically-inac- tive, US-positive 9.9± 4.4 SD Clinically- inactive, US negative
11.6±3.8 Median 11.9
(IQR 7.3-15.19) Median 10.7 (8.9-12.5)
ILAR JIA categories
(%) 46 (52) persistent
oligoarticular 15 (17) extended oligoarticular 15 (17) RF neg polyarticular 12 (14) other
14 (40) persistent oligoarticular 12 (34.3) extended oligoarticular 9 (25.7) polyarticular
18 (46.1) persistent oligoarticular 12 (30.8) extended oligoarticular 4 (10.3) RF neg polyarticular 2 (5.1) ERA 2 (5.1) Psoriatic 1 (2.6) Systemic
10 (25) persistent oligoarticular 12 (30) extended oligoarticular J 15 (38) RF neg polyarticular 1 (3) RF positive polyarticular 2 (5) ERA N clinically inactive patients
(N patients off treatment) / N clinically inactive joints scanned at baseline
88 (28) / 3872
(44 joints per patient) 35 (9) / 3298
total assessed joints 24 (13) / 2028
total assessed joints 40 (5) / 289 total assessed joints
Duration of
clinical inactive disease at baseline
0.9 years±0.6 in US-positives 1.9±0.8 years in US-negatives
1.9±2.2 years minimum 3 months 1 year (0.5-1.8)
Observation period 4 years 30 months 2 years Not reported
(median 22 months) N patients clinically inactive
US positive at baseline N joints clinically inactive US-positive at baseline
20/88 (22%) patients 38 joints (0.98%)
24/35 (68.6%) patients 122 joints (3.7%)
14/39 (35.9%) patients 45 joints (2,2%)
18/40 (45%) patients 24 joints (8.3%) Clinically-inactive,
US-positive patients/joints at baseline that flared (N, %)
15/20 (75%) patients NA
NA
18/122 (14.7%) joints
15/39 (38.5%) 17/45 (37.7%) joints
NA
4/24 (16.6%) joints Interval of relapse At 1 year 45% of US
positive patients and 6% of US negative relapse
Mean 5.95 months
(0.25-18) Median 10.6 months
(6.3-13.7) Median 12 months (3-24)
Healthy controls Enrolled Not enrolled Enrolled Not enrolled
Risk of flare in clinically inactive US positive patients
Increased
(with predictive value higher in presence of positive PD signal)
Increased only in presence of positive PD signal
Not increased Not increased
N, number; PD, power Doppler; JIA, juvenile idiopathic arthritis; US, ultrasound; NA, not assigned; RF, rheumatoid factor; ERA, early rheumatoid arthritis
pared to the US-negative group. In the paper by De Lucia et al, however, the risk was higher at patient level but not at individual joint level. Of note, healthy controls were recruited only in two of the studies [18,20], one of which also included clinically active joints as positive controls [18].
Discussion
The use of US in paediatric rheumatology has gained wide attention, but its real role in clinical practice is still a matter of debate [22]. While its usefulness in detect- ing synovitis in difficult to examine joints such as the ankle and as guidance in arthrocentesis is now estab- lished [3,23], studies on other possible uses have yielded conflicting results [20,21]. It is claimed that it can detect subclinical synovitis [23-25], but without a gold stand- ard such as histology or MRI the superiority of US when compared to clinical examination cannot be confirmed even if the experience maturated in RA of the adult [26]
pushes us forward in this direction.
The possibility to predict disease flares with a non- invasive, radiation-free method such as US would be on the other hand very useful for the clinician, who in case of disease remission has few elements to judge when to taper medical treatment. Studies in this regard have been performed in RA [27,28], but data in JIA are scanty.
We performed a systematic review on this topic and could identify only 4 articles which met our inclusion criteria. Of these, 2 were in favour of an increased risk and two were not. Several factors could influence this discrepancy. Series included were not large, with a total of less than 200 patients altogether and 3 of the 4 articles including <40 cases. With regard to ILAR classification, the majority of cases were in the persistent oligoarticular JIA category, which is the most common, and the over- all distribution reflects the epidemiology of the diseases
also considering the ethnicity of the different series [29].
Studies were all prospective and in the only one which included more than one centre [18], the procedures were standardized. Of note, healthy controls are particularly important when scanning joints of a growing child and only two of the studies [18,20] included a control group.
In one study not only negative but also positive controls were included [18]; this also would have been desirable since the presence of active synovitis can sometimes be difficult to distinguish from physiological vascularization in an immature joint [1,3]. Of note in the two works with a control group, the percentage of patients and joints with US detected subclinical synovitis at baseline were sensi- bly lower than the other (see table II).
The predictive value of PD signal in articular carti- lage of JIA patients should be considered with caution due to the possibility of its misinterpretation, being this data also observed in healthy, growing children. Discord- ant results regarding this aspect also appear in the four articles analyzed in this metanalysis. In fact, De Lucia et al [18] observed an increased predictive value for re- lapse when grey scale findings were associated with PD abnormalities, Miotto et al [19] found the presence of PD signal essential for predicting the risk of flare, Zhao et al [21], documented abnormal PD signal only in a low percentage of children (without a correlation with disease flare), and Magni Manzoni et al [20], surprisingly found a greater frequency of PD signal in patients with persis- tent inactive disease with respect to those who flared.
Among the different factors which could influence the risk of relapse, duration of inactive disease and du- ration of follow-up varied among studies, which would also be possible bias in interpreting the results. Variations in medical treatment during the observation period was accounted for in the article by De Lucia et al [18], but not in the others. This is obviously another bias, since the risk of relapse can be influenced by the tapering or withdrawal of drug therapy.
While in RA studies have demonstrated that subclini- cal inflammation can be found on US in patients which are in clinical remission according to the clinical meas- ures [30] and that US-detected residual synovitis can pre- dict the risk of relapse and structural progression [31,32], in paediatrics this has not been definitely proven. Indeed, despite growing evidence supporting the potential role of US in the monitoring of patients, the use of US is still a matter of debate even in RA. Two recent RCTs (TASER and ARTIC) have demonstrated that a treatment strategy based on the US assessment did not lead to an improved clinical outcome in comparison with a conventional treat to target (T2T) approach, suggesting that the systematic use of US in the follow-up of RA patients would be not Fig 1. PRISMA flow chart
justified [11,12]. In children many pitfalls linked to the structure, anatomy, and physiology of the growing joint make this type of study difficult to perform and interpret.
The real question of whether US-detected subclinical synovitis is a risk factor for relapse, according to our study, cannot be answered both for the paucity of data and for discordance of results in the few articles included in the review.
First of all, studies which include a gold standard (if histology is not possible, at least MRI) would be neces- sary to judge the superiority of US when compared to clinical examination; second, a standardization of meth- odology is required d and is still not widespread; and third, appropriate positive and negative control groups are necessary in order to bypass as much as possible the interpretation biases that are so common in this regard.
Conflict of interest: none References
1. Collado P, Vojinovic J, Nieto JC, et al. Toward Standard- ized Musculoskeletal Ultrasound in Pediatric Rheumatol- ogy: Normal Age-Related Ultrasound Findings. Arthritis Care Res (Hoboken) 2016;68:348-356.
2. Windschall D, Collado P, Vojinovic J, et al. Age-Related Vascularization and Ossification of Joints in Children: An International Pilot Study to Test Multiobserver Ultrasound Reliability. Arthritis Care Res (Hoboken) 2020;72:498- 3. Magni-Manzoni S. Ultrasound in juvenile idiopathic arthri-506.
tis. Pediatr Rheumatol Online J 2016;14:33.
4. Cipolletta E, Filippucci E, Incorvaia A, et al. Ultrasound- Guided Procedures in Rheumatology Daily Practice: Fea- sibility, Accuracy, and Safety Issues. J Clin Rheumatol 2021;27:226-231.
5. Nordberg LB, Lillegraven S, Aga AB, et al. The Impact of Ultrasound on the Use and Efficacy of Intraarticular Glu- cocorticoid Injections in Early Rheumatoid Arthritis: Sec- ondary Analyses From a Randomized Trial Examining the Benefit of Ultrasound in a Clinical Tight Control Regimen.
Arthritis Rheumatol 2018;70:1192-1199.
6. Wakefield RJ, Green MJ, Marzo-Ortega H, et al. Should ol- igoarthritis be reclassified? Ultrasound reveals a high prev- alence of subclinical disease. Ann Rheum Dis 2004;63:382- 7. Naredo E, Bonilla G, Gamero F, Uson J, Carmona L, Laffon 385.
A. Assessment of inflammatory activity in rheumatoid ar- thritis: a comparative study of clinical evaluation with grey scale and power Doppler ultrasonography. Ann Rheum Dis 2005;64:375-381.
8. Brown AK, Quinn MA, Karim Z, et al. Presence of signifi- cant synovitis in rheumatoid arthritis patients with disease- modifying antirheumatic drug-induced clinical remission:
evidence from an imaging study may explain structural pro- gression. Arthritis Rheum 2006;54:3761-3773.
9. Brown AK, Conaghan PG, Karim Z, et al. An explanation for the apparent dissociation between clinical remission and continued structural deterioration in rheumatoid arthritis.
Arthritis Rheum 2008;58:2958-2967.
10. Foltz V, Gandjbakhch F, Etchepare F, et al. Power Doppler ultrasound, but not low-field magnetic resonance imaging, predicts relapse and radiographic disease progression in rheumatoid arthritis patients with low levels of disease ac- tivity. Arthritis Rheum 2012;64:67-76.
11. Haavardsholm EA, Aga AB, Olsen IC, et al. Ultrasound in management of rheumatoid arthritis: ARCTIC randomised controlled strategy trial. BMJ 2016;354:i4205.
12. Dale J, Stirling A, Zhang R, et al. Targeting ultrasound remission in early rheumatoid arthritis: the results of the TaSER study, a randomised clinical trial. Ann Rheum Dis 2016;75:1043-1050.
13. Caporali R, Smolen JS. Back to the future: forget ultra- sound and focus on clinical assessment in rheumatoid ar- thritis management. Ann Rheum Dis 2018;77:18-20.
14. Halyabar O, Mehta J, Ringold S, Rumsey DG, Horton DB.
Treatment Withdrawal Following Remission in Juvenile Idiopathic Arthritis: A Systematic Review of the Literature.
Paediatr Drugs 2019;21:469-492.
15. Moher D, Liberati A, Tetzlaff J, Altman DG; PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Int J Surg 2010;8:336-341.
16. Uman LS, Chambers CT, McGrath PJ, Kisely S, Mat- thews D, Hayton K. Assessing the quality of randomized controlled trials examining psychological interventions for pediatric procedural pain: recommendations for quality im- provement. J Pediatr Psychol 2010;35:693-703.
17. Jadad AR, Moore RA, Carroll D, et al. Assessing the qual- ity of reports of randomized clinical trials: is blinding nec- essary? Control Clin Trials 1996;17:1-12.
18. De Lucia O, Ravagnani V, Pregnolato F, et al. Baseline ultrasound examination as possible predictor of relapse in patients affected by juvenile idiopathic arthritis (JIA). Ann Rheum Dis 2018;77:1426-1431.
19. Miotto E Silva VB, Mitraud SAV, Furtado RNV, Natour J, Len CA, Terreri MTSELRA. Patients with juvenile idi- opathic arthritis in clinical remission with positive power Doppler signal in joint ultrasonography have an increased rate of clinical flare: a prospective study. Pediatr Rheumatol Online J 2017;15:80.
20. Magni-Manzoni S, Scire CA, Ravelli A, et al. Ultra- sound-detected synovial abnormalities are frequent in clinically inactive juvenile idiopathic arthritis, but do not predict a flare of synovitis. Ann Rheum Dis 2013;72:223- 21. 228.Zhao Y, Rascoff NE, Iyer RS, et al. Flares of Disease in Children with Clinically Inactive Juvenile Idiopathic Ar- thritis Were Not Correlated with Ultrasound Findings. J Rheumatol 2018;45:851-857.
22. Cimaz R, Giani T, Caporali R. What is the real role of ul- trasound in the management of juvenile idiopathic arthritis?
Ann Rheum Dis 2020;79:437-439.
23. Roth J. Emergence of Musculoskeletal Ultrasound Use in Pediatric Rheumatology. Cur Rheumatol Rep 2020;22:14.
24. Rebollo-Polo M, Koujok K, Weisser C, Jurencak R, Bruns A, Roth J. Ultrasound findings on patients with juvenile idiopathic arthritis in clinical remission. Arthritis Care Res (Hoboken) 2011;63:1013-1019.
25. Haslam KE, McCann LJ, Wyatt S, Wakefield RJ. The detec- tion of subclinical synovitis by ultrasound in oligoarticular juvenile idiopathic arthritis: a pilot study. Rheumatology (Oxford) 2010;49:123-127.
26. Anandarajah A, Thiele R, Giampoli E, et al. Patients with rheumatoid arthritis in clinical remission manifest persis- tent joint inflammation on histology and imaging studies. J Rheumatol 2014;41:2153-2160.
27. Zufferey P, Scherer A, Nissen MJ, et al. Can Ultrasound Be Used to Predict Loss of Remission in Patients with RA in a Real-life Setting? A Multicenter Cohort Study. J Rheumatol 2018;45:887-894.
28. Filippou G, Sakellariou G, Scire CA, et al. The predictive role of ultrasound-detected tenosynovitis and joint synovi- tis for flare in patients with rheumatoid arthritis in stable
remission. Results of an Italian multicentre study of the Italian Society for Rheumatology Group for Ultrasound:
the STARTER study. Ann Rheum Dis 2018;77:1283-1289.
29. Consolaro A, Giancane G, Alongi A, et al. Phenotypic vari- ability and disparities in treatment and outcomes of child- hood arthritis throughout the world: an observational cohort study. Lancet Child Adolesc Health 2019;3:255-263.
30. Di Matteo A, Mankia K, Azukizawa M, Wakefield RJ. The Role of Musculoskeletal Ultrasound in the Rheumatoid Ar- thritis Continuum. Curr Rheumatol Rep 2020;22:41.
31. Nguyen H, Ruyssen-Witrand A, Gandjbakhch F, Constantin A, Foltz V, Cantagrel A. Prevalence of ultrasound-detected residual synovitis and risk of relapse and structural progres- sion in rheumatoid arthritis patients in clinical remission: a systematic review and meta-analysis. Rheumatology (Ox- ford) 2014;53:2110-2118.
32. Han J, Geng Y, Deng X, Zhang Z. Subclinical Synovitis As- sessed by Ultrasound Predicts Flare and Progressive Bone Erosion in Rheumatoid Arthritis Patients with Clinical Re- mission: A Systematic Review and Metaanalysis. J Rheu- matol 2016;43:2010-2018.
