Objective To investigate the assessed value of tumor-associated macrophages ( TAMs ) and KIT expression for liver metastasis in pancreatic neuroendocrine tumors (PNETs) and patients′outcome. Methods A total of 79 patients who underwent surgical resection and pathologically diagnosed as PNETs in the Department of Hepatopancreatobiliary Surgery in Sun Yat-sen Memorial Hospital from January 1995 to May 2015 were enrolled.The immunohistochemical staining of CD68 and KIT were detected and the correlations with clinicopathological factors were analyzed.Results Of 79 PNETs cases, CD68 and KIT in tumor tissue were overexpressed in 30(38%) and 35(44.3%) cases, respectively.CD68 overexpression was associated with tumor infiltration ( P<0.001 ), AJCC stage 7 ( P<0.001 ), liver metastasis ( P<0.001 ) and early recurrence (P=0.019).Patients with low CD68 level had significantly better survival than those with high CD68 expression ( P=0.0002 ).KIT overexpression was correlated with WHO 2010 and AJCC stage 7 (P<0.001;P=0.002), nonfunctional status of the tumor (P=0.002) and liver metastasis (P=0.026). The survival period of patients with low KIT expression was greatly longer than those with high KIT level (P=0.0013).CD68 and KIT co-overexpression was observed in patients with tumor invasion (P<0.001), advanced WHO and AJCC stage (both P<0.001) and better prognostic survival (P=0.0057).Multivariate analysis showed that CD68 overexpression (HR:2.9;95%CI:1.16~7.23;P=0.033) was an independent prognostic factor for PNETs.Conclusions CD68 and KIT overexpression is correlated with advanced disease stage, higher risk for liver metastasis and worse survival.CD68 is an independent prognostic factor for PNETs.

Clustering is a prevalent analytical means to analyze single cell RNA sequencing (scRNA-seq) data but the rapidly expanding data volume can make this process computationally challenging. New methods for both accurate and efficient clustering are of pressing need. Here we proposed Spearman subsampling-clustering-classification (SSCC), a new clustering framework based on random projection and feature construction, for large-scale scRNA-seq data. SSCC greatly improves clustering accuracy, robustness, and computational efficacy for various state-of-the-art algorithms benchmarked on multiple real datasets. On a dataset with 68,578 human blood cells, SSCC achieved 20% improvement for clustering accuracy and 50-fold acceleration, but only consumed 66% memory usage, compared to the widelyused software package SC3. Compared to k-means, the accuracy improvement of SSCC can reach 3-fold. An R implementation of SSCC is available at https://github.com/Japrin/sscClust.
Algorithms ; Animals ; Cluster Analysis ; Computational Biology ; methods ; Databases as Topic ; Gene Expression Profiling ; methods ; Humans ; Mice ; Sequence Analysis, RNA ; Single-Cell Analysis ; Software ; Statistics, Nonparametric