To the Editor:

Weber et al¹ reported results of a study on the clinical significance of genes coamplified with MYCN in neuroblastoma tumors. According to univariate Kaplan-Meier survival analysis on a series of 98 MYCN amplified neuroblastomas, DDX1 coamplification was found to identify a subgroup of advanced stage neuroblastoma with a more favorable outcome (P = .027, log-rank test). To exclude a possible confounding effect of incomplete follow-up, a subsequent analysis was limited to 51 patients with a follow-up survival of at least 6 years (as it was shown that no statistically relevant change in survival probability occurred after this follow-up period). This analysis revealed a more significantly improved prognosis for patients with DDX1 coamplification compared with patients with MYCN amplification alone (P = .01, log-rank test). In order to enforce their findings, the authors refer to our study² stating that the survival curves show a trend (without statistical significance) toward a better prognosis for patients presenting with a DDX1 coamplified neuroblastoma tumor. Weber et al attributed this lack of statistical significance to the shorter follow-up times and smaller cohort of patients with MYCN amplified tumors (n = 66).

As the observation of improved survival associated with DDX1 coamplification may have important consequences for therapy stratification, independent confirmation is required. Therefore, we re-evaluated our data using updated survival information for the patients. In addition to the 66 patients reported,² we included seven more MYCN-amplified tumors (from the Cancer Research Unit, Newcastle, UK). Of the 73 patients with MYCN amplification, 39 showed DDX1 coamplification (mean and median follow-up time of all survivors, 73.62 and 65.73 months, respectively). Kaplan-Meier survival analysis revealed no significant differences in event-free survival (P = .069, log-rank test) or overall survival (P = .090, log-rank test). In analogy with Weber et al, we subsequently excluded 10 patients with a follow-up time of less than 6 years for Kaplan-Meier survival analysis. Even now, the survival data of the remaining 63 patients did not reveal significant differences in event-free survival (P = .074, log-rank test; relative risk, 0.64; 95% CI, 0.38 to 1.09; Fig 1A) or overall survival (P = .11; relative risk, 0.66; 95% CI, 0.39 to 1.13; Fig 1B) of DDX1 coamplified neuroblastomas versus MYCN-only amplified tumors. Likewise, the trend in the data from Weber et al toward a better prognosis for MYCN amplified neuroblastoma patients with NAG coamplification (P = .09, log-rank test) could not be confirmed in our data set (63 patients: event free survival, P = .84; overall survival, P = .97, log-rank test).

View larger version:

• In this window
• In a new window

• PowerPoint Slide for Teaching

Fig 1.

Kaplan-Meier event-free survival (A) and overall survival (B) analyses for DDX1 coamplification in tumors with a long-term follow-up of more than 6 years after initial diagnosis (n = 63).

Summarizing, our data indicate that—in contrast to the observations by Weber et al—DDX1 coamplification has no significant prognostic (adverse or beneficial) value in MYCN amplified tumors. Furthermore, our own data, previous studies,³ and the observations by Weber et al show that the gene coamplification frequency within the MYCN amplicon correlates with the physical distance to MYCN. Hence, these findings strongly suggest that the process of coamplification occurs coincidentally, and is not subject to selection, further supporting our conclusions. Further arguments against a beneficial effect of DDX1 coamplification come from published reports attributing oncogenic characteristics to DDX1.³

In conclusion, the absence of significant prognostic power of DDX1 coamplification after re-evaluation of a large series of MYCN-amplified tumors from patients with long follow-up times, the marginally significant P values obtained in Weber et al (0.027 and 0.01), the physical association between gene coamplification and distance to the core target of amplification, and the oncogenic properties of DDX1 strongly suggest that the observations made by Weber et al constitute at the best a marginal effect. Other large cohorts of uniformly treated neuroblastoma tumors need to be studied to clarify these discrepant findings and to demonstrate whether DDX1 coamplification has any favorable predictive power. The question remains though how such an effect would be translated to the clinic in order to achieve an improved outcome for both patient groups.

Authors' Disclosures of Potential Conflicts of Interest

The authors indicated no potential conflicts of interest.