Óscar González-Velasco

Purpose: Neuroblastoma, a common cancer in children, is characterized by striking clinical heterogeneity, ranging from spontaneous regression to fatal disease. Precise risk classification is therefore essential for appropriate treatment stratification. Given that genotoxic chemotherapy constitutes the mainstay of neuroblastoma treatment, it is particularly important to avoid overtreatment in patients with biologically favorable disease, notably in the heterogeneous group of current non–high-risk patients. Here, we tested the hypothesis that telomere maintenance mechanisms (TMM), which have been proposed as key hallmark of high-risk neuroblastoma, may improve neuroblastoma risk classification over current standard of care. Methods: We examined the prognostic accuracy of TMM in comparison with current standard-of-care prognostic variables in 366 patients with neuroblastoma, including 259 non–high-risk patients (test set, n = 116; validation set, n = 143). Patient event-free survival (EFS) and overall survival (OS) were determined according to TMM and currently used molecular classification by chromosomal copy-number alterations or gene expression profiles. The prognostic impact of the three molecular classifications was compared pairwise and with standard-of-care prognostic variables in multivariable analyses. Results: TMM-based and gene expression-based classification accurately distinguished non–high-risk patients with excellent and adverse EFS and OS in both the test and validation sets. On multivariable analyses of standard-of-care prognostic factors, TMM-based classification remained the only risk factor in the final models (EFS, hazard ratio [HR], 6.4 [95% CI, 3.9 to 10.6], P < .001; OS, HR, 29.4 [95% CI, 9.6 to 90.0], P < .001), in contrast to the two other molecular classifiers. In pairwise comparisons, TMM-based classification outperformed the two other molecular classifications in models built on both EFS and OS ( P < .001 each). Conclusion: TMM-based classification improves risk assessment in neuroblastoma, providing a rationale for clinical implementation as a new standard for risk stratification of patients with neuroblastoma.

Supplementary Table 7 shows the prevalence of alterations in all genes assayed for the five most commonly occurring EML4-ALK variants in the NSCLC cohort (V3, V1, V2, V5' and V5).

Supplementary Figure 8 shows histological analysis and quantification of cancer type marker gene expression and signaling protein phosphorylation in EML4-ALK V1 and V3 tumors of different tumor suppressor genotypes

Supplementary Figure 4 shows data on the adaptively sample mean size of tumors with each tumor suppressor gene inactivation at the sgRNA-level in the Lenti-sgTSG19-sgV1/Cre and Lenti-sgTSG19-sgV3/Cre pools in both Cas9EGFP and Cas9-negative control mice.

Supplementary Figure 5 shows the schematic of Lenti-sgTSG75-sgV1/Cre and Lenti-sgTSG75-sgV3/Cre pools, tumor initiation in Cas9-negative mice to determine the representation of Lenti-sgRNA/Cre vectors, and additional data from broader screen of tumor suppressor function in V1- and V3-driven lung cancer including tumor burden, tumor number and effect.

Supplementary Table 11 shows the impact of genes (Score RTN and significance) on EML4-ALK-V3 tumorigenesis in the context of lorlatinib treatment.

Supplementary Figure 6 show data on the adaptively sample mean size of tumors with each tumor suppressor gene inactivation at the sgRNA-level in the Lenti-sgTSG75-sgV1/Cre and Lenti-sgTSG75-sgV3/Cre pools in both Cas9EGFP and Cas9-negative control mice.

Supplementary Figure 2 show the increase tumorigenesis of V3 tumors in a different mouse model, the general cloning strategy to generate the triple sgRNA vectors, and the efficiency of GFP inactivation in EML4-Alk-driven autochthonous tumors.

Supplementary Table 4 shows adaptively sampled mean (ASM) and adaptively sampled 95th percentile size (AS95) and significance of genes in the Lenti-sgTSG75 pools on EML4-ALK-V1 and V3-driven lung tumorigenesis.

Supplementary Table 5 shows the information of the sgRNAs included in the Tuba-seqUltra pool using in the in vivo lorlatinib response experiment and their overlap with sgRNAs in the Lenti-sgTSG19 and Lenti-sgTSG75 pools.