de Graaf P. Diagnostic Imaging in Retinoblastoma. Amsterdam: Vrije Universiteit 2012, 256 pages.
Diagnostic Imaging in Retinoblastoma is a concise review of retinoblastoma as it applies to the neuroradiologist. Author Pim de Graaf, who is currently completing his training in neuroradiology at the VU University Medical Center in Amsterdam, has organized his discussion into essentially six sections. The first section is a brief yet comprehensive 10-page review of the epidemiology, genetics, pathology, dissemination pathways, clinical presentation, staging, and treatment of retinoblastoma. The second section is an even more succinct 4-page overview of the diagnostic imaging modalities available. The author discusses the historical utility of CT scan and its replacement by MRI as its imaging technology has overcome the early drawbacks of lack of spatial resolution and low signal-to-noise ratio. Although ophthalmoscopy and sonography remain the initial diagnostic tools, the author advises that MRI may aid diagnosis of indeterminate presentations of leukocoria, and that MRI should be employed in all patients with retinoblastoma to evaluate local tumor extension, brain involvement and CSF metastases, and associated midline PNETs . The next three sections are a collection of the author’s nine research publications, which delve more deeply into the details of MR imaging in retinoblastoma as it relates to value in differential diagnosis, detection of disease extent, and specific evaluation of angiogenesis and tumor vitality. Finally, the author summarizes in just 10 pages the diagnostic imaging options, value in differential diagnosis, detection of tumor extent, evaluation of angiogenesis and tumor vitality, clinical implications, and the future of imaging in retinoblastoma.
MR imaging characteristics of the tumor may guide therapeutic decisions. For example, abnormal contrast enhancement in the anterior eye segment may be a marker of a more aggressive tumor due to the morphologically and pathophysiologically abnormal neovessels of iris angiogenesis. Additionally, MR imaging of tumor necrosis may indicate a more aggressive tumor that has outgrown its vascular supply. Dynamic contrast-enhanced MRI exploits the richness of vascular supply and vascular endothelial permeability of tumor neovascularity via analysis of contrast uptake in the tumor tissue through imaging before, during, and after contrast administration. Early phase enhancement correlates with microvessel density, while late enhancement correlates with necrosis. Diffusion-weighted imaging with apparent diffusion coefficient maps may also provide valuable information regarding the increased restriction of free diffusion through a more highly cellular viable tumor environment or decreased restriction of free diffusion through disrupted cell membranes of necrotic tumor cells.
The first research publication presented (de Graaf et al, Pediatric Radiology, 2012) discusses guidelines for imaging retinoblastoma including imaging principles, MRI protocol standardization, and radiation exposure, and is invaluable for the practicing neuroradiologist. de Graaf includes a funduscopic photograph and sonographic image along with multiple well-captioned MR images to provide the radiologist with a complete appreciation of all the imaging available to the ophthalmologist.
The next publication (de Graaf et al, Radiology, 2007) is a retrospective design over 12 years evaluating eye size, from which the authors conclude that all imaging measurements were significantly smaller in eyes with retinoblastoma than in normal eyes, and that eye size decreases as tumor burden increases.
The next article (de Graaf et al, American Journal of Neuroradiology, 2007), is a case report of MR imaging findings of retinal dysplasia, which mimics intraocular tumors. The authors include funduscopic images as well as gross and microscopic photographs in addition to the MR images.
The following three publications concern disease extent. The first (de Graaf et al, Radiology, 2005) is a retrospective study of fifty-eight eyes with retinoblastoma that were evaluated with T1, T2, and gadolinium-enhanced T1 weighted imaging and correlated with histopathological diagnosis. Sensitivity and specificity for choroidal invasion was 73% and 72%, respectively. Sensitivity and specificity for prelaminar optic nerve invasion was 66% and 96%, respectively. Sensitivity and specificity for postlaminar optic nerve invasion was 50% and 100%, respectively. This publication may be the most useful for the practicing radiologist due to the detailed descriptions of MR imaging findings with correlative annotated MR and histopathological images. The next publication (de Graaf et al, British Journal of Ophthalmology, 2006) is a case report of optic nerve enhancement which persisted after chemotherapy treatment, prompting enucleation. Histopathologic findings included inflammation and endothelial proliferation, however there was no tumor invasion of the optic nerve, and the post-treatment MRI was falsely positive. The third publication (Rodjan et al, American Journal of Neuroradiology, 2010) details associated brain abnormalities including pineoblastoma, pineal cysts, and structural brain abnormalities such as corpus callosum agenesis, Dandy-Walker variant, dilated ventricles, vermis hypoplasia, and trignocephalia, all of which the radiologist must be aware.
The author has compiled nine concise publications regarding the imaging characteristics of retinoblastoma, which are of great importance to the radiologist, replete with funduscopic, histopathologic, gross anatomic, and MR images. His thorough yet succinct introduction and summary complete the fund of knowledge necessary for the radiologist to participate as an important member of the medical team treating patients with retinoblastoma.
