To estimate the diagnostic accuracy of post gadolinium high resolution Magnetic Resonance Imaging (MRI) for early diagnosis, optic nerve invasion and detection of intraocular calcification in retinoblastoma using European Retinoblastoma Imaging Collaboration (ERIC) guidelines.

Abstract

PURPOSE: To estimate the diagnostic accuracy of post gadolinium high resolution magnetic resonance images for early diagnosis, optic nerve invasion and detection of intraocular calcification in retinoblastoma using European Retinoblastoma Imaging Collaboration (ERIC) guidelines.

SUBJECT AND METHOD:

A total of forty five patients were included in this study all were less than 12 years of age and referred for the MRI of suspected retinoblastoma. Standard MRI technique of orbits and brain with contrast were obtained according to ERIC guidelines by Magneton Avanto 1.5 Tesla (SIEMENS) scanner at Dr. Ziauddin University Hospital Karachi. The results of MRI were compared with histopathological results obtained after surgery (enucleation) of the patients.

results:

Mean age of the patients was 19.98 + 18.41 months. Sex distribution shows, 25 patients (55.6%) were male while remaining 20 patients (44.4%) were female. Duration of Symptoms was 33.6 + 30.5 days. True positive cases were 39 and true negative cases were 4. Diagnostic accuracy of CE-MRI for the detection of RB was found to be 95.5% (43/45) [p value 0.0001], sensitivity and specificity was found to be 97.5% (32/32) and 80% (4/5) respectively. [p value 0.0001]. For the detection of calcification high resolution MRI shows 82.2% accuracy while US/CT shows 84.4% accuracy for the detection of calcification. Optic nerve invasion was diagnosed with an accuracy of 55.5%. Out of 45 eyes, 30 eyes show retinal detachment which was diagnosed on MRI with 77.8% accuracy. 40% of RB shows exophytic growth pattern and 37.8% shows endophytic tumor growth while 11.1% showed mixed exo and endophytic growth pattern.

CONCLUSION:

High resolution MRI has superior contrast and spatial resolution compared to computed tomograph (CT) or other available imaging techniques. MRI is highly sensitive for the diagnosis of retinoblastoma as well as for the detection of optic nerve invasion which may be helpful in the treatment planning. MRI cannot substitute CT in detecting tumour calcification in retinoblastoma but with increasing experience and new techniques it should be possible to renounce CT scans in the majority of cases of retinoblastoma.

key words:

Retinoblastoma, ERIC, Diagnostic Accuracy, MRI