Anaplastic lymphoma kinase (ALK) is a receptor tyrosine kinase predominantly expressed in the developing nervous system. Recently, mutated ALK has been identified as a major oncogene associated with familial and sporadic neuroblastomas (NBL). Yet, a direct correlation between endogenous expression level of the ALK protein, oncogenic potential, and clinical outcome has not been established. We investigated ALK genetic mutations, protein expression/phosphorylation, and functional inhibition both in NBL-derived cell lines and in 34 localized and 48 advanced/metastatic NBL patients. ALK constitutive phosphorylation/activation was observed in high-ALK expressing cells, harboring either a mutated or a wild-type receptor. No activation was found in cell lines with low expression of wild-type ALK. After 72 hours of treatments, small molecule ALK inhibitor CEP-14083 (60 nmol/L) induced growth arrest and cell death in NBL cells overexpressing wild-type (viability: ALK^high 12.8%, ALK^low 73%, P = 0.0035; cell death: ALK^high 56.4%, ALK^low 16.2%, P = 0.0001) or mutated ALK. ALK protein expression was significantly up-regulated in advanced/metastatic compared with localized NBLs (ALK overexpressing patients: stage 1-2, 23.5%; stage 3-4, 77%; P < 0.0001). Interestingly, protein levels did not always correlate with ALK genetic alterations and/or mRNA abundance. Both mutated and wild-type ALK receptor can exert oncogenic activity in NBL cells. However, wild-type ALK receptor requires a critical threshold of expression to achieve oncogenic activation. Overexpression of either mutated or wild-type ALK defines poor prognosis patients. Alternative mechanisms other than direct mutations and/or gene amplification regulate the ALK level of expression in NBL cells. Wild-type ALK is a potential therapeutic target for advanced/metastatic NBLs. [Cancer Res 2009;69(18):7338–46]

Several neuroblastoma (NB) susceptibility loci have been identified within LINC00340, BARD1, LMO1, DUSP12, HSD17B12, DDX4, IL31RA, HACE1 and LIN28B by genome-wide association (GWA) studies including European American individuals. To validate and comprehensively evaluate the impact of the identified NB variants on disease risk and phenotype, we analyzed 16 single nucleotide polymorphisms (SNPs) in an Italian population (370 cases and 809 controls). We assessed their regulatory activity on gene expression in lymphoblastoid (LCLs) and NB cell lines. We evaluated the cumulative effect of the independent loci on NB risk and high-risk phenotype development in Italian and European American (1627 cases and 2575 controls) populations. All NB susceptibility genes replicated in the Italian dataset except for DDX4 and IL31RA, and the most significant SNP was rs6435862 in BARD1 (P = 8.4×10^–15). BARD1 showed an additional and independent SNP association (rs7585356). This variant influenced BARD1 mRNA expression in LCLs and NB cell lines. No evidence of epistasis among the NB-associated variants was detected, whereas a cumulative effect of risk variants on NB risk (European Americans: P _trend = 6.9×10^–30, Italians: P _trend = 8.55×10¹³) and development of high-risk phenotype (European Americans: P _trend = 6.9×10^–13, Italians: P _trend = 2.2×10^–1) was observed in a dose-dependent manner. These results provide further evidence that the risk loci identified in GWA studies contribute to NB susceptibility in distinct populations and strengthen the role of BARD1 as major genetic contributor to NB risk. This study shows that even in the absence of interaction the combination of several low-penetrance alleles has potential to distinguish subgroups of patients at different risks of developing NB.

Radiomics is defined as the use of automated or semi-automated post-processing and analysis of multiple features derived from imaging exams. Extracted features might generate models able to predict the molecular profile of solid tumors. The aim of this study was to develop a predictive algorithm to define the mutational status of EGFR in treatment-naïve patients with advanced non–small cell lung cancer (NSCLC). CT scans from 109 treatment-naïve patients with NSCLC (21 EGFR-mutant and 88 EGFR-wild type) underwent radiomics analysis to develop a machine learning model able to recognize EGFR-mutant from EGFR-WT patients via CT scans. A “test–retest” approach was used to identify stable radiomics features. The accuracy of the model was tested on an external validation set from another institution and on a dataset from the Cancer Imaging Archive (TCIA). The machine learning model that considered both radiomic and clinical features (gender and smoking status) reached a diagnostic accuracy of 88.1% in our dataset with an AUC at the ROC curve of 0.85, whereas the accuracy values in the datasets from TCIA and the external institution were 76.6% and 83.3%, respectively. Furthermore, 17 distinct radiomics features detected at baseline CT scan were associated with subsequent development of T790M during treatment with an EGFR inhibitor. In conclusion, our machine learning model was able to identify EGFR-mutant patients in multiple validation sets with globally good accuracy, especially after data optimization. More comprehensive training sets might result in further improvement of radiomics-based algorithms. Significance: These findings demonstrate that data normalization and “test–retest” methods might improve the performance of machine learning models on radiomics images and increase their reliability when used on external validation datasets.

Non-small cell lung cancer is one leading cause of death worldwide, and patients would greatly benefit from an early diagnosis. Since targeted and immunotherapies have emerged as novel approaches for more tailored treatments, repeated assessments of the tumor biology have become pivotal to drive clinical decisions. Currently, tumor tissue biopsy is the gold standard to investigate potentially actionable biomarkers, but this procedure is invasive and may prove inadequate to represent the whole malignancy. In this regard, liquid biopsy represents a minimally invasive and more comprehensive option for early detection and investigation of this tumor. Today, cell-free DNA is the only approved circulating marker to select patients for a targeted therapy. Conversely, the other tumor-derived markers (i.e., circulating tumor cells, miRNAs, exosomes, and tumor educated platelets) are still at a pre-clinical phase, although they show promising results for their application in screening programs or as prognostic/predictive biomarkers. The main challenges for their clinical translation are the lack of reliable cutoffs and, especially for miRNAs, the great variability among the studies. Moreover, no established tool has been approved for circulating tumor cells and exosome isolation. Finally, large prospective clinical trials are mandatory to provide evidence of their clinical utility.

Exome and deep sequencing of clinically aggressive neuroblastoma reveal somatic mutations that affect key pathways involved in cancer progression