Clinical evaluation of ultra high field MRI for three-dimensional visualization of tumour size in uveal melanoma patients, with direct relevance to treatment planning.
Jan-Willem Beenakker1, Teresa Ferreira1, Karina Soemarwoto1, Lorna Grech Fonk1, Stijn Genders1, Wouter Teeuwisse1, Andrew Webb1, and Gregorius Luyten1

1Leiden University Medical Centre, Leiden, Netherlands


Recent advances in ocular MRI make it possible to acquire high resolution three dimensional images of uveal melanoma in eye tumour patients, allowing a much better assessment of the maximal tumour prominence compared to conventional clinical ultrasound measurements. Nine uveal melanoma patients were examined on a 7 Tesla using a custom-built eye-coil. Eye-motion artefacts were minimized by the use of a cued-blinking protocol. For all patients the MR-images showed a slightly lower tumour prominence. For two of these patients this resulted in a substantial change in treatment planning, saving an eye that would otherwise have been removed.


Uveal melanoma is the most common primary intraocular malignant tumour in adults. In the past, surgical removal of the eye was the main treatment, but in the last ten years various eye- and vision-saving treatments have become available. The optimal treatment is mainly determined by the location and size of the tumour. Ruthenium plaque brachytherapy, for example, spares most of the healthy surrounding tissue, resulting in the optimal preservation of visual function. The total radiation dose delivered by a ruthenium plaque decreases as the distance from the plaque increases, therefore this treatment is only effective if the tumour prominence - the distance between the outside of the sclera, where the plaque is located, and the tumour - is below 7 mm. At present, the extent of the tumour is primarily determined using 2D ultrasound. To measure the prominence of the tumour correctly, the transducer needs to be positioned perpendicular to the tumour, which is often not possible due to the surrounding tissue, e.g. the nose. These oblique cuts through the tumour potentially result in an over-estimation of the tumour size, with the degree of over-estimation being patient/tumour specific meaning that no simple correction algorithm can be applied.Recent advances in ocular MRI make it possible to image the complete tumour in three dimensions 1-3, potentially allowing a much better determination of the maximal tumour prominence.1 In this work, we have estimated the tumour dimensions in uveal melanoma patients at 7 Tesla ocular MRI, and compare these values with conventional ultrasound imaging in order to provide improved information for treatment options.


Nine uveal melanoma patients were examined on a Philips Achieva 7 Tesla whole-body magnet (Best, The Netherlands) using a custom-built eye-coil shown in Figure 1. Eye-motion artefacts were minimized by the use of a cued-blinking protocol consisting of a regular break every 3 seconds, in which the scanner was automatically paused and the subjects were visually instructed to blink. MR-images were acquired using a 3D inversion recovery turbo gradient echo technique (MP-RAGE) with an inversion time of 1280 ms, a shot interval of 3 s, and a turbo field echo factor of 92; the TR/TE/tip angle were: 6.7 ms/3.4 ms/16º. The scan time was 3 minutes and resulted in a spatial resolution of 0.4 x 0.4 x 0.9 mm3. The ophthalmic ultrasound measurements were made using a B-scan ultrasonography (Quantel Medical Aviso, Cournon d'Auvergne France) with a transducer frequency of 10 MHz, an exploration angle of 50º, focus of 24–26 mm, axial resolution of 200 µm and lateral resolution of 600 µm (data supplied by the manufacturer).


The MRI protocols resulted in high-resolution motion-free images, figure 2, of the eye in which the tumour and surrounding tissues could clearly be discriminated. For each subject the maximal prominence of the tumour was assessed using a three-dimensional viewer. Figure 3 shows results from two specific patients, comparing the ultrasound and MR images. In one of these patients the ultrasound maximum prominence of 7.5 mm was reduced to 6.2 mm from the MR images, thus changing the clinical treatment plan from eye removal to ruthenium plaque placement. In general, the MR-images showed a 0.5 to 2 mm (average 1.3 mm) smaller tumor prominence than the ultrasound values, which can be attributed to the oblique cuts through the tumour made by the ultrasound.


For all patients the MR-images showed a slightly lower tumour prominence. For two of these patients this resulted in a substantial change in treatment plant. The original ultrasound measurements showed a tumour that was slightly too large for ruthenium plaque therapy, meaning that the eye would be removed. The MRI, however, showed a slightly smaller tumour that would still be eligible for ruthenium plaque therapy, which meant that the eye could be spared. Due to the uncertainties of the ultrasound, especially the potential oblique cuts through the tumour, the final decision was based on the MR-images, and ruthenium plaque therapy was offered. In addition to the described diagnostic value of the MRI scans, the evaluation of these patients revealed two other potential benefits of MRI compared to ultrasound. Firstly, the full three-dimensional data on the tumour geometry can be used for a more precise planning for the different forms of external beam radiotherapy of uveal melanoma: this is currently based on two-dimensional fundus photography and ultrasound images. Secondly, the high intra-tumour contrast compared to ultrasound, as shown in figure 3, potentially allows for a non-invasive classification of the tumour, which could replace the current invasive praxis of intra-ocular biopsies which has significant risks.


High-field ocular MRI gives a more accurate measurement of the tumour dimensions than conventional ultrasound, which can result in significant changes in the prescribed treatment. As a result of this study, high-field ocular MR-imaging has been included in the standard clinical care for uveal melanoma patients within our hospital.


This study was supported by the Landelijke Stichting Blinden en Slechtzienden, Utrecht, The Netherlands.


1. Beenakker, J. W. M., van Rijn, G. A., Luyten, G. P. M. & Webb, A. G. High-resolution MRI of uveal melanoma using a microcoil phased array at 7 T. NMR Biomed. 26, 1864–1869 (2013).

2. Graessl A, Muhle M, Schwerter M, et al. Ophthalmic magnetic resonance imaging at 7 T using a 6 channel transceiver radiofrequency coil array in healthy subjects and patients with intraocular masses. Invest Radiol 49:260–270 (2014).

3. Paul K, Graessl A, Rieger J, et al. Diffusion-sensitized ophthalmic magnetic resonance imaging free of geometric distortion at 3.0 and 7.0 T: a feasibility study in healthy subjects and patients with intraocular masses. Invest Radiol 50:309–321 (2015).


Figure 1. Schematics and pictures of the receive eye coil, with an LC-balun. Mirrors were attached to the coil-­housing to allow the patient to look at a screen positioned at the end of the magnet bore.

Figure 2. High-­resolution images of patient 1 (A,B,C) and patient 2 (D,E,F). The tumour prominence (green line) was measured on the images acquired with the 3D MPRAGE sequence (A,D). (B,E) 2D spin-­echo acquisitions measured perpendicular to the tumour. (C) A 3D spoiled gradient FFE shows a hypointense structure within the tumour. (D) A fat-­ suppressed inversion recovery sequence can be used to screen for tumour infiltrations 2 through the sclera.

Figure 3. Comparison of ultrasonic and MRI-based measurements of maximum tumour prominence.

Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)