Diagnostic accuracy in retinoblastoma.

This study was done to ascertain the accuracy of diagnosis and possible risk of misdiagnosis in suspected retinoblastoma cases by different pre-operative diagnostic procedures like fundoscopy, ultrasonography, computerised tomography, and aqueous cytology study. Twenty-four eyes excised with diagnosis of suspected retinoblastoma were included in the study. Final diagnosis was reached by histopathological examination of excised eyes and compared with pre-operative diagnosis by different diagnostic procedures and accuracy of each procedure was calculated. Twenty-two cases were finally diagnosed as retinoblastoma and 2 cases (8.33%) as other than retinoblastoma. Computerised tomography and aqueous humour cytology were two most specific (100%) tests. Computerised tomography was the most sensitive (95.4%) test. Aqueous cytology was poorly sensitive (9.09%). Both ultrasonography and fundoscopy demonstrated sensitivity over 90% but with low specificity (50%). When test results of fundoscopy and ultrasonography were combined, specificity of this combination was 100% and sensitivity was 95.4%, which match exactly the values of computerised tomography. Combination of fundoscopy and ultrasonography has accuracy as high as that of computerised tomography. In patients with clear media this combination can be used as an effective alternative to computerised tomography to reach a diagnosis with comparable accuracy.

A cross-sectional study on aqueous humour lactate dehydrogenase level in retinoblastoma.

Lactate dehydrogenase enzyme is a non-specific marker found raised in different body fluids in various pathological conditions. This cross-sectional study was done in a medical college of Kolkata, India to find out lactate dehydrogenase level of aqueous humour in retinoblastoma cases. Aqueous humour lactate dehydrogenase level from 21 histology proved retinoblastoma cases and 21 controls was estimated and compared. Range of lactate dehydrogenase levels in retinoblastoma cases were 863-4819 IU/l; mean being 3396.9 IU/l; and in controls range was 48-146 IU/l; mean being 90.6 IU/l; 76.19% of retinoblastoma cases had value over 3000 IU/l. None of the controls had value above 150 IU/l. The difference between lactate dehydrogenase values of cases and controls was significant (p<0.001).

An observational study on retinoblastoma cases attending a medical college in Calcutta.

Excision of the eyeball was done in 23 cases with diagnosis of retinoblastoma in a medical college of Calcutta. Of them, 21 cases were histologically proved to be retinoblastoma. A clinicopathological study of these 21 cases was done. Retinoblastoma patients were between 5.5 and 71 months of age, 57.1% were females, in 90.4% cases there was unilateral involvement and left eye was affected in 68.4% patients. All the patients had negative family history. Majority of the patients (85.7%) presented with leucocoria. Multiple presenting features were present in 5 patients. Aqueous humour cytology revealed malignant cells in 9.5% patients. Imprint cytology of cut end of optic nerve stump was negative for malignant cells in all the cases. Histopathological study revealed presence of Flexner-Wintersteiner rosettes in 23.8%, pseudorosettes in 28.6%, anterior chamber involvement in 38%, choroid invasion in 42.8%, optic nerve involvement in 28.6%, necrosis in 28.6% and calcification in 9.5% cases. After one year of follow-up, none developed any metastasis except one patient who had developed intracranial metastasis.