| BackgroundGastric cancer(GC)is one of the malignancies with the highest morbidity and mortality in the world.At present,the prognostic model of GC is mainly based on the Tumor-Node-Metastasis(TNM)staging system.However,due to the existence of individual heterogeneity,the prognosis of patients with similar TNM stage and receiving the same treatment varies greatly.Ubiquitinproteasome system(UPS)participates in a wide range of biological processes,including cell growth,cell cycle,DNA metabolism and signal transduction.Previous studies have shown that its important subunit PSMD1 is associated with many human cancers.PSMD1 is highly expressed in undifferentiated thyroid cancer,breast cancer and other tumor tissue.Our preliminary study found that PSMD1 expression in GC tissue was higher than that in corresponding normal tissue.The purpose of this study was to explore the expression of PSMD1 in GC tissue and to evaluate its prognostic value.Materials and MethodsTotal RNA extraction,reverse transcription polymerase chain reaction and realtime quantitative PCR were performed on 36 fresh frozen GC tissue samples to study the expression of PSMD1 in tumor tissue.From June 2006 to May 2009,the expression of PSMD1 in the paraffin-embedded tumor and normal tissue from the training cohort(n=241)and the paraffin-embedded tumor tissue from the validation cohort(n=170)were retrospectively analyzed by immunohistochemistry(IHC).PSMD1 expression in each specimen was evaluated by determining a semi-quantitative H score,which was calculated by multiplying the result of a 4-stepscale(0=negative,0.5=weak staining,1=moderate staining,1.5=strong staining)and the fraction of positively stained cells that ranged from 0 to 100%.The H-scores for PSMD1 staining were dichotomized at the median to classify GC patients into high and low PSMD1 expression groups.The clinical characteristics,pathologic results and long-term outcomes were reviewed.The Kaplan-Meier survival curve and univariate and multivariate Cox regression model were used to evaluate the prognostic value of PSMD1 expression.Based on the expression of PSMD1 and other risk factors related to GC,we generated two nomograms by R software to evaluate the prognosis of patients with gastric cancer.Furthermore,the nomograms were calibrated and verified,and their clinical efficacies were evaluated.ResultsIn the real-time quantitative PCR experiment,compared with non-tumor tissue,PSMD1 expression in most GC intra-tumoral tissue was up-regulated on mRNA level(25/36;69.4%).In the IHC experiment of the training cohort(n=241),the median H score of the PSMD1 expression were 43.0(Range:6.0-111.0)in the non-tumor tissue and 63.0(8.0-141.0)in the tumor tissue.The PSMD1 expression in tumor tissue was higher than that in the non-tumor tissue(P<0.0001).Therefore,the high expression of PSMD1 was stable in GC tissue.The percentage of patients with high intra-tumoral PSMD1 expression increased moderately from TNM stages Ⅰ to Ⅳ(stage Ⅰ:28.1%,Ⅱ:32.7%,Ⅲ:56.8%,Ⅳ:69%).The expression of PSMD1 was much higher in advanced stage GC(H score,stages Ⅰ-Ⅱ(n=80)vs.stages Ⅲ-Ⅳ(n=161):35.5(8.0130.0)vs.78.0(18.0-141.0),P=0.0006),suggesting its association with disease progression.Kaplan-Meier survival curve analysis showed that PSMD1 expression in nontumor tissue was not associated with DFS and OS in the training cohort.However,patients with high intra-tumoral PSMD1 expression were associated with lower 5-year DFS and OS than patients with low intra-tumoral PSMD1 expression(5-year DFS:17.5%vs.42.1%;hazard ratio(HR)=1.977(95%CI:1.428-2.739);5-year OS:24.2%vs.48.9%;HR=1.992(1.417-2.801);all P<0.001).We obtained similar results in the validation cohort(high vs.low PSMD1 expression groups,5-year DFS:HR=1.781(1.171-2.708),P=0.006;5-year OS:HR=1.751(1.144-2.682),P=0.008),which confirmed the poor prognostic significance of high PSMD1 expression in GC patients.According to the subgroup analysis of factors,such as TNM stage,the degree of differentiation,CEA and CA199,the DFS and OS in the high PSMD1 expression group were almost worse than those in the low expression group(P<0.05).Univariate and multivariate Cox proportional hazard analysis demonstrated that PSMD1 was an independent prognostic factor for predicting DFS(HR=1.603(1.1312.271),P=0.008)and OS(HR=1.584(1.096-2.289),P=0.014)in GC patients.The two nomograms that were developed by integrating PSMD1 expression and the TNM staging system showed better prediction of 1-year,3-year and 5-year DFS and OS than TNM staging system alone.Recalibration was not required because the calibration curves for the two nomograms did not deviate from the reference line.In the training cohort,the C-indexes for DFS and OS prediction were 0.708(95%CI:0.670-0.746)and 0.712(95%CI:0.671-0.752),respectively.In the validation cohort,the C-indexes for DFS and OS prediction were 0.704(0.651-0.757)and 0.711(0.656-0.767),respectively.The calibration curves yielded good agreement between the predicted and observed outcomes for both DFS and OS.Furthermore,the performances of the nomograms for DFS and OS in both the training and validation cohorts were superior relative to the TNM.staging system alone.The C-indexes of TNM stage for DFS and OS prediction were 0.681(0.648-0.715)and 0.678(0.640-0.716)in the training cohort,and 0.660(0.610-0.709)and 0.661(0.610-0.712)in the validation cohort,respectively.Using the X-tile,the composite scoring was divided into three risk groups that accurately discriminated between patients with good,intermediate,and poor prognosis(DFS,P<0.0001;OS,P<0.0001).The cutoff scores for DFS were 98 and 166 and for OS were 103 and 175,respectively.ConclusionsPSMD1 is stably and highly expressed in GC tissue.GC patients with high expression of PSMD1 in intra-tumoral tissue had worse DFS and OS than those with low expression.The results of multivariate Cox proportional hazard analysis showed that the high expression of PSMD1 was an independent poor predictor of the prognosis of gastric cancer patients.PSMD1 expression and TNM staging system comprehensive analysis can make nomogram prediction model visualization,which could be applied to calculate the patient’s survival probability.Integrated analysis of PSMD1 expression with TNM staging system provides a better prognosis prediction model for GC patients and shows potential to aid both clinicians and patients in regard to counseling,individualized adjuvant treatment decision-making and follow-up scheduling. |
PDF Full Text Request