Comparison of the prognostic significance of the Shimada classification and N-myc copy number in neuroblastoma.

This report outlines the experience with the first 20 patients (8 males and 12 females) enrolled in the Canadian National Neuroblastoma Diagnostic Laboratory, The study population ranged in age at diagnosis from one month to 11 years. Fourteen children had advanced (stage 3 or 4) disease. Tumors were sampled extensively and were classified, at the time of accession, according to the 'Shimada' histopathological scheme. A portion of each tumor was analyzed for N-myc oncogene copy number. Nine tumors were classified as having 'favourable' histopathological features and 11 as 'unfavourable'. N-myc oncogene amplification, of 3 or more copies, was found in 2 of 9 tumors with 'favourable' histology and 5 of 11 with 'unfavourable' features. The follow-up interval was at least two years from initial diagnosis. The Shimada classification was more accurate than the N-myc oncogene copy number (p<0.01) in predicting clinical outcome. The sensitivity and specificity for Shimada histopathological classification were 100% and 92% respectively, while corresponding values were 75% and 42% for N-myc copy number. Our experience indicates that, when assessing prognosis in neuroblastoma, Shimada classification performs better than the N-myc copy number.