The etiology of neuroendocrine tumors remains poorly defined. Although neuroendocrine tumors are in some cases associated with inherited genetic syndromes, such syndromes are rare. The majority of neuroendocrine tumors are thought to be sporadic. We performed a genome-wide association study (GWAS) to identify potential genetic risk factors for sporadic neuroendocrine tumors. Using germline DNA from blood specimens, we genotyped 909,622 SNPs using the Affymetrix 6.0 GeneChip, in a cohort comprising 832 neuroendocrine tumor cases from Dana-Farber Cancer Institute and Massachusetts General Hospital and 4542 controls from the Harvard School of Public Health. An additional 241 controls from Dana-Farber Cancer Institute were used for quality control. We assessed risk associations in the overall cohort, and in neuroendocrine tumor subgroups. We identified no potential risk associations in the cohort overall. In the small intestine neuroendocrine tumor subgroup, comprising 293 cases, we identified risk associations with three SNPs on chromosome 12, all in strong LD. The three SNPs are located upstream of ELK3, a transcription factor implicated in angiogenesis. We did not identify clear risk associations in the bronchial or pancreatic neuroendocrine subgroups. This large-scale study provides initial evidence that presumed sporadic small intestine neuroendocrine tumors may have a genetic etiology. Our results provide a basis for further exploring the role of genes implicated in this analysis, and for replication studies to confirm the observed associations. Additional studies to evaluate potential genetic risk factors for sporadic pancreatic and bronchial neuroendocrine tumors are warranted.

In this article, we study finite mixtures of linear mixed‐effects (FMLME) models that are useful for longitudinal regression modelling in the presence of heterogeneity in both fixed and random effects. These models are computationally challenging when the number of covariates is large, and traditional variable selection techniques become expensive to implement. We introduce a penalized likelihood approach, and propose a nested EM algorithm for efficient numerical computations. The resulting estimators are shown to possess consistency and sparsity properties, and to be asymptotically normally distributed. We illustrate the performance of our method through simulations and a real data example. The Canadian Journal of Statistics 41: 596–616; 2013 © 2013 Statistical Society of Canada

No abstract

No abstract