Sensitive detection of numerical and structural aberrations of chromosome 1 in neuroblastoma by interphase fluorescence in situ hybridization. Comparison with restriction fragment length polymorphism and conventional cytogenetic analyses.

Chromosome I abnormalities are indicators of prognosis in neuroblastoma (NB) but are not yet routinely exploited because conventional methods are technically demanding. We evaluated the pertinence of interphase cytogenetics fluorescence in situ hybridization (FISH) for the analysis of chromosome I in NB, compared with conventional methods. Deletion of Ip was detected in 8 of 9 cell lines analyzed by both FISH and restriction fragment length polymorphism (RFLP), but was evidenced in only 2 cases by conventional cytogenetics, painting analysis being required to reveal the other cases. The chromosome I number evaluated by FISH reflected the total chromosome modal number obtained by cytogenetics. Twenty-eight specimens obtained from ultrasound-guided punctures, surgical biopsies of the primary tumor and bone-marrow aspirates were studied by FISH on frozen cytocentrifuged smears; 12 had a chromosome I trisomy and 16 a disomy. Requirements for a reliable control analysis of Ip deletion by RFLP were met in only 23 cases. The retention of 2 alleles was observed in 15 cases and Ip deletion in 7, by both techniques. In one case, an interstitial deletion of Ip was evidenced only by RFLP, and one of 5 cases analyzed only by FISH had a Ip deletion. Although FISH might be improved by using additional probes, it presents major advantages for routine exploitation. Determining Ip deletion in individual cells makes it possible to analyze small and heterogeneous tumoral specimens; the technique requires only a few hours and can easily be standardized in non-specialized laboratories. The number of chromosome I homologues per cell might serve as a rapid screening for ploidy.

[CD44 expression: a new prognostic factor in neuroblastoma]. / Expression de CD44: un nouveau facteur pronostique pour le neuroblastome.

CD44 gene products are potential markers of aggressiveness in different tumor models, a result which prompted us to study clinical neuroblastoma (NB) specimens. CD44 expression was determined by immunostaining of 52 NB with a monoclonal antibody (J173) directed against an epitope common to all CD44 isoforms. All tumors were from patients (pts) with newly diagnosed NB treated with standardized protocols. They were classified according to international criteria [11]. CD44 immunoreactivity was detected in 37 tumors (71%). CD44 was expressed in 100% of favorable NB stages (1, 2 or 4S), but only 50% of advanced NB (stages 3 and 4) (p = 0.0001), suggesting that the absence rather that the overexpression of CD44 is a signal of tumor aggressiveness. The cumulative event-free survival was significantly longer in pts with CD44-positive tumors as compared to pts with CD44-negative tumors (p < 10(-5)). More importantly, progression-free survival was also significantly higher in CD44-positive pts within the high-risk group (p < 0.01). In univariate analyses, we tested the prognostic value of tumor expression of CD44 in comparison with tumor stage, age, tumor histology and presence or absence of N-myc proto-oncogene amplification. All five measures had significant prognostic value. The expression of CD44 and the absence of N-myc amplification were the most powerful predictors of a favorable clinical outcome. In a multivariate analysis of these measures, CD44 expression and tumor stage were the only independent prognostic factors for the prediction of patient survival. NB is the first clinical model described in which tumor aggressiveness correlates with a repression rather than a stimulation of CD44 expression. We recommend the use of CD44 as an additional biological marker in the initial staging of neuroblastoma.

Evaluation of CD44 prognostic value in neuroblastoma: comparison with the other prognostic factors.

CD44 gene products are potential markers of aggressiveness in different tumour models, a result which prompted us to study clinical neuroblastoma (NB) specimens. CD44 expression was determined by immunostaining of 52 tumour samples from newly diagnosed NB with a monoclonal antibody (J173) directed against an epitope common to all CD44 isoforms. CD44 immunoreactivity was detected in 37 of the tumours (71%). CD44 was expressed in all 22 NBs with favourable prognoses (stages 1, 2 or 4S), but only 50% (15/30) of advanced NB (stages 3 and 4) (P < 10(-4)), suggesting that the absence, rather than the overexpression, of CD44 is a signal of tumour aggressiveness. The cumulative progression-free survival was significantly longer in patients with CD44 positive tumours compared with patients with CD44 negative tumours (P < 10(-5)). More importantly, progression-free survival was also significantly higher in CD44 positive patients within the high-risk group (P < 0.01). In univariate analysis, we tested the prognostic value of tumour expression of CD44 in comparison with tumour stage, age, tumour histology, and presence or absence of amplification of the MYCN protooncogene. All five measures had significant prognostic value. The expression of CD44 and the absence of MYCN amplification were the most powerful predictors of a favourable outcome. In a multivariate analysis of these measures, CD44 expression and tumour stage were the only independent prognostic factors for the prediction of patient survival. NB is the first clinical model described in which tumour aggressiveness correlates with repression rather than stimulation of CD44 expression. We recommend the use of CD44 as an additional biological marker in the initial staging of NB.