MR imaging characteristics of uveal melanoma with histopathological validation

Abstract

Purpose To evaluate the magnetic resonance imaging (MRI) characteristics of uveal melanoma (UM), to compare them with fundoscopy and ultrasound (US), and to validate them with histopathology. Methods MR images from 42 UM were compared with US and fundoscopy, and on 14 enucleated cases with histopathology. Results A significant relationship between the signal intensity on T1 and pigmentation on histopathology was found (p=0.024). T1 hyperintense UM were always moderately or strongly pigmented on histopathology, while T1-hypointense UM were either pigmented or non-pigmented. Mean apparent diffusion coefficient (ADC) of the UM was 1.16 ± 0.26 × 10−3 mm2/s. Two-thirds of the UM had a wash-out and the remaining a plateau perfusion time-intensity curve (TIC). MRI was limited in evaluating the basal diameter of flat tumors. US tends to show larger tumor prominence (0.5mm larger, p=0.008) and largest basal diameter (1.4mm larger, p<0.001). MRI was good in diagnosing ciliary body involvement, extrascleral extension, and optic nerve invasion, but limited on identifying scleral invasion. An increase of tumor prominence was associated with lower ADC values (p=0.030) and favored a wash-out TIC (p=0.028). An increase of tumor ADC correlated with a plateau TIC (p=0.011). Conclusions The anatomical and functional MRI characteristics of UM were comprehensively assessed. Knowing the MRI characteristics of UM is important in order to confirm the diagnosis and to differentiate UM from other intra-ocular lesions and because it has implications for treatment planning. MRI is a good technique to evaluate UM, being only limited in case of flat tumors or on identifying scleral invasion.

Histopathologic and MR Imaging Appearance of Spontaneous and Radiation-Induced Necrosis in Uveal Melanomas: Initial Results

“…The mean ADC of ocular melanomas we measured in the present cohort, in both the irradiated and the control group (0.78 × 10 −3 mm 2 /s and 0.92 × 10 −3 mm 2 /s respectively) is lower than that reported in a recently published study by Ferreira et al [ 41 ] but is only partially inconsistent with this latter. In our opinion such a low ADC value found in the control group could probably be related to our study cohort (large tumors needing enucleation); it is, in fact, closer to that reported by Erb-Eigner et al [ 61 ] in whose study uveal melanomas had a mean diameter of 14 mm (very similar to our).…”

“…In our opinion such a low ADC value found in the control group could probably be related to our study cohort (large tumors needing enucleation); it is, in fact, closer to that reported by Erb-Eigner et al [ 61 ] in whose study uveal melanomas had a mean diameter of 14 mm (very similar to our). In line with this, also Ferreira et al found uveal melanomas with larger tumor prominence having lower ADC values [ 41 ]. The even lower ADC value we found in the irradiated group could be associated with tumor aggressiveness since these uveal melanomas represent tumor regrowth (recurrence), but honestly, the number of available measurements (only four) is too small to make any assumptions.…”

“…We performed MR examinations through a 1.5 T scanner and a neurovascular array coil. The use and availability of MRI systems operating at higher magnetic field strengths (≥3 T) coupled with dedicated receive surface coils undoubtedly ensure higher spatial resolution and signal-to-noise ratio [ 13 , 41 , 62 , 63 ]. Moreover, although our MR protocol included DWI sequences, it lacks other functional sequences, such as dynamic contrast-enhanced (DCE)-MRI and perfusion weighted-imaging (PWI) that have proven to be useful in the differential diagnosis of intraocular tumors and in evaluating tumor response to radiotherapy and could have a promising value in detecting lesions with a more severe prognosis [ 13 , 41 , 63 , 64 ].…”

“…The MR appearance of tumor necrosis in both the irradiated and nonirradiated groups was evaluated on T1-weighted, T2-weighted, and diffusion-weighted imaging (DWI) acquisitions. As previously described by Ferreira et al, the signal intensity of necrotic areas on T1- and T2-weighted sequences was classified as hyper-, iso-, or hypointense, taking as a reference the signal intensity of the choroid and of the extrinsic eye muscles on T1- and T2-weighted images respectively [ 41 ]. In addition, the radiologists performed quantitative measurements of the apparent diffusion coefficient (ADC) of the tumors, as previously described [ 18 ], through a representative region of interest (ROI), taking care to exclude the most peripheral parts of the lesion, distortion artifacts, and macroscopically evident necrotic foci.…”

“…During the past decades, some studies have been carried out comparing an irradiated with a nonirradiated control group in order to differentiate the effects of radiotherapy from spontaneous degenerative changes, including necrosis, in uveal melanomas [ 36 , 37 , 38 , 39 ]; nevertheless, this topic has been investigated from the histopathological point of view, and a radiologic-pathologic correlation was not performed. Only in the last few years de Jong et al and Ferreira et al correlated MR findings with histopathology in retinoblastomas and uveal melanomas respectively [ 40 , 41 ]. However, in regard to both spontaneous and radiation-induced necrosis, some aspects are not yet fully understood, either in terms of pathogenetic mechanism or radiologic imaging.…”