Predictive value of a serum tumor biomarkers scoring system for clinical stage Ⅱ/Ⅲ rectal cancer with neoadjuvant chemoradiotherapy

BACKGROUND Multiple classes of molecular biomarkers have been studied as potential predictors for rectal cancer(RC)response.Carcinoembryonic antigen(CEA)is the most widely used blood-based marker of RC and has proven to be an effective predictive marker.Cancer antigen 19-9(CA19-9)is another tumor biomarker used for RC diagnosis and postoperative monitoring,as well as monitoring of the therapeutic effect.Using a panel of tumor markers for RC outcome prediction is a practical approach.AIM To assess the predictive effect of pre-neoadjuvant chemoradiotherapy(NCRT)CEA and CA19-9 levels on the prognosis of stage II/III RC patients.METHODS CEA and CA19-9 levels were evaluated 1 wk before NCRT.According to the receiver operating characteristic curve analysis,the optimal cut-off point of CEA and CA19-9 levels for the prognosis were 3.55 and 19.01,respectively.The novel serum tumor biomarker(NSTB)scores were as follows:score 0:Pre-NCRT CEA<3.55 and CA19-9<19.01;score 2:Pre-NCRT CEA>3.55 and CA19-9>19.01;score 1:Other situations.Pathological information was recorded according to histopathological reports after the operation.RESULTS In the univariate analysis,pre-NCRT CEA<3.55[P=0.025 for overall survival(OS),P=0.019 for disease-free survival(DFS)],pre-NCRT CA19-9<19.01(P=0.014 for OS,P=0.009 for DFS),a lower NSTB score(0-1 vs 2,P=0.009 for OS,P=0.005 for DFS)could predict a better prognosis.However,in the multivariate analysis,only a lower NSTB score(0-1 vs 2;for OS,HR=0.485,95%CI:0.251-0.940,P=0.032;for DFS,HR=0.453,95%CI:0.234-0.877,P=0.019)and higher pathological grade,node and metastasis stage(0-I vs II-III;for OS,HR=0.363,95%CI:0.158-0.837,P=0.017;for DFS,HR=0.342,95%CI:0.149-0.786,P=0.012)were independent predictive factors.CONCLUSION The combination of post-NCRT CEA and CA19-9 was a predictive factor for clinical stage II/III RC patients receiving NCRT,and the combined index had a stronger predictive effect.

Face recognition based on subset selection via metric learning on manifold

With the development of face recognition using sparse representation based classification(SRC), many relevant methods have been proposed and investigated. However, when the dictionary is large and the representation is sparse, only a small proportion of the elements contributes to the l1-minimization. Under this observation,several approaches have been developed to carry out an efficient element selection procedure before SRC. In this paper, we employ a metric learning approach which helps find the active elements correctly by taking into account the interclass/intraclass relationship and manifold structure of face images. After the metric has been learned, a neighborhood graph is constructed in the projected space. A fast marching algorithm is used to rapidly select the subset from the graph, and SRC is implemented for classification. Experimental results show that our method achieves promising performance and significant efficiency enhancement.