Fluorescence optical imaging and 3T-MRI for detection of synovitis in patients with rheumatoid arthritis in comparison to a composite standard of reference

Abstract

Objectives

To address whether Indocyanine Green (ICG) enhanced fluorescence optical imaging (FOI) is more sensitive than magnetic resonance imaging (MRI) in the detection of synovitis of the wrist and finger joints in rheumatoid arthritis and to analyze the performance of FOI depending on the grade of synovitis.

Methods

Twenty patients with highly active rheumatoid arthritis (mean DAS28-ESR 5.25 ± 1.0) and thirteen healthy volunteers underwent clinical examination, FOI and contrast-enhanced 3T-MRI. Joints were rated by three independent readers semiquantitatively (grade 0–3: no, low, moderate and high grade synovitis) and compared to a semiquantitative composite standard of reference (cSOR, grade 0–3) that incorporated clinical parameters, FOI and MRI results.

Results

2.868 evaluations in 956 joints were performed. FOI had an overall sensitivity of 57.3% and a specificity of 92.1%, whereas MRI had a sensitivity of 89.2% and a specificity of 92.6%. The sensitivity of FOI increased with the degree of synovitis to 65.0% for moderate and severe synovitis (specificity 88.1%) and 76,3% for severe synovitis (specificity 80.5%). The performance of FOI decreased with the degree of synovitis with false negative results predominantly for mild (156/343, 45.5%) and moderate (160/343, 46.6%) synovitis and false positive FOI evaluations predominantly based on weak (grade 1) signals (133/163, 81,6%).

Conclusion

FOI has a lower sensitivity than 3T-MRI in the detection of synovitis of the hand and finger joints. The diagnostic performance of FOI decreases with the degree of synovitis and with the strength of FOI signals.

Introduction

Rheumatoid arthritis is a chronic progressive disease that, if untreated, leads to joint destruction in most cases [1]. Active synovitis results in bone erosions very early in the course of disease, placing irreversible damage to the joints, which in turn impairs functional outcome in the long-term [2], [3], [4]. Early and effective therapy may slow or even halt disease progression, but there may be substantial joint damage caused by ongoing synovitis despite clinical remission [5], [6]. Therefore, an accurate detection of joint inflammation is one of the most important challenges in the diagnosis and treatment of rheumatoid arthritis (rA).

Currently, different imaging modalities are applied in rA, each having specific advantages and diasadvantages. Conventional radiography is used as a baseline and follow-up examination in rA, but detects bone destruction in a time delayed manner restricting its use in therapy guidance. In contrast, ultrasonography (US) in greyscale mode (GSUS) and in power doppler mode (PDUS) reliably detect synovitis very early in the disease course and help to predict radiographic structural damage [7], [8], [9], [10], [11], [12]. However, despite efforts in the standardization of the method, US remains highly operator- and device-dependent and is restricted due to its expenditure of time. Magnetic resonance imaging (MRI) offers the advantage to sensitively detect inflammation along with precise information on soft tissue and bones and has become a standard of reference in the assessment of inflammatory joint disease. Furthermore, the degree of synovitis and bone marrow oedema detected by MRI have been shown to be independent predictors of joint destruction in rA [13], [14], [15]. However, the use of MRI is limited by costs and restricted availability.

Recently, fluorescence optical imaging (FOI) has been proposed as a new method for the detection of inflammation in the wrist and the finger joints [16], [17], [18], [19], [20]. The principle of FOI is the detection of increased perfusion at sites of synovitis with the help of the fluorophore indocyanine green (ICG) which is exposed to near-infrared light. A previous study reported a sensitivity of FOI of 76% at a specificity of 54% in the detection of synovitis in patients with established rheumatoid arthritis with 1.5T MRI as a standard of reference [16]. In untreated patients with early rheumatoid arthritis a sensitivity of 86% and a specificity of 63% were reported in another study [19]. In both studies, a higher rate of positive signals was reported with FOI than with MRI. The authors therefore supposed FOI to detect subclinical inflammation and to be potentially superior to 1.5T MRI in the detection of joint inflammation. However, a recent study reported a lower overall sensitivity of 40% and a specificity of 85% for FOI in the detection of synovitis compared to 3T MRI. Since the sensitivity of FOI was depending on the degree of inflammation of synovitis seen on MRI, concerns over the diagnostic accuracy of FOI especially in patients with moderate or low grade joint inflammation were raised [17].

Our study prospectively compared the diagnostic performance of ICG-enhanced FOI and MRI to detect synovitis in the hands of patients with moderately to highly active rheumatoid arthritis depending on the grade of synovitis using a composite reference standard derived from FOI, 3T MRI and clinical data.