Y022

Real-world use of non-echoplanar diffusion weighted MRI imaging for detection and clinical decision-making in Graves’ orbitopathy

Feeney C¹, Nagendran S¹, Lee V¹, Lingam RK², Frahman F¹

¹ Eye Department, Central Middlesex Hospital, London, UK

² Department of Radiology, Northwick Park & Central Middlesex Hospital, London Northwest University Healthcare NHS Trust, London, UK

Background

The Clinical Activity Score (CAS) is often used in clinical practice to guide treatment decisions in Graves’ orbitopathy (GO) but has limitations, particularly in its potential to evaluate posterior orbital disease. We have previously demonstrated that non-echoplanar DWI can be a useful adjunct to clinical assessment in GO and compares well to other standard MRI modalities.^1,2 The main objectives of this study were to i) correlate DWI imaging with CAS and ii) determine the clinical decisions and outcomes influenced by our radiological evaluation.

Methods

This was a retrospective observational study of 31/88 patients who had at least one DWI scan (60 in total) and corresponding CAS (same observer-VL) as part of routine clinical care in a multidisciplinary GO clinic from 2012 to 2017. For each scan, an experienced Head and Neck Radiologist (RL) derived an apparent diffusion coefficient (ADC) for each extra-ocular muscle (EOM). Spearman’s rank correlation coefficient was used to determine the relationship between CAS and ADC. A Decision Tree was constructed to evaluate clinical decisions taken on the basis of DWI activity. Receiver-operator curves (ROC) were plotted for mild GO and dysthyroid optic neuropathy (DON).

Results

There was a significant positive correlation between CAS and ADC (n=368 EOMs) (r_s=0.403 CI 0.312-0.489, P<0.0001). In addition, ADC values were significantly higher in the CAS ≥3 group compared to the CAS <3 group, P<0.0001. Our Decision Tree identified a third ‘intermediate’ severity cohort where activity on DWI was particularly useful in making clinical decisions. ADC performed well as a diagnostic test in predicting DON (AUC 0.974 95% CI 0.93-1.0). An exploratory secondary analysis suggested that the right orbit was more commonly involved than the left with a predilection for the right inferior rectus and relative sparing of the lateral recti.

Conclusions

Non-echoplanar DWI correlates well with CAS in our patients and was a useful adjunct to CAS in making clinical decisions especially in patients with ‘intermediate’ severity GO. DWI may be useful in identifying patients at risk of DON and may be a useful research tool to test hypotheses regarding temporal and spatial patterns of disease activity in GO.

1.           Lingam RK, Mundada P, Lee V. Novel use of non-echo-planar diffusion weighted MRI in monitoring disease activity and treatment response in active Grave's orbitopathy: An initial observational cohort study. Orbit (Amsterdam, Netherlands). 2018;37(5):325-330.

2.           Ritchie AE, Lee V, Feeney C, Lingam RK. Using Nonechoplanar Diffusion-weighted MRI to Assess Treatment Response in Active Graves Orbitopathy: A Novel Approach with 2 Case Reports. Ophthalmic plastic and reconstructive surgery. 2016;32(3):e67-70.