Dataset Information

A novel gene-pair signature for relapse-free survival prediction in colon cancer.

ABSTRACT: Background:Colon cancer (CC) patients with early relapse usually have a poor prognosis. In this study, we aimed to identify a novel signature to improve the prediction of relapse-free survival (RFS) in CC. Methods:Four microarray datasets were merged into a training set (n=1,045), and one RNA-sequencing dataset was used as a validation set (n=384). In the training set, microarray meta-analysis screened out 596 common RFS-related genes across datasets, which were used to construct 177,310 gene pairs. Then, the LASSO penalized generalized linear model identified 16 RFS-related gene pairs, and a risk score was calculated for each sample according to the model coefficients. Results:The risk score demonstrated a good ability in predicting RFS (area under the curve [AUC] at 5 years: 0.724; concordance index [C-index]: 0.642, 95% CI: 0.615-0.669). High-risk patients showed a poorer prognosis than low-risk patients (HR: 3.519, 95% CI: 2.870-4.314). Subgroup analysis reached consistent results when considering multiple confounders. In the validation set, the risk score had a similar performance (AUC at 5 years: 0.697; C-index: 0.696, 95% CI: 0.627-0.766; HR: 2.926, 95% CI: 1.892-4.527). When compared with a 13-gene signature, a 15-gene signature, and TNM stage, the score showed a better performance (P<0.0001; P=0.0004; P=0.0125), especially for the patients with a longer follow-up (R2=0.988, P<0.0001). When the follow-up was >5 years (n=314), the score demonstrated an excellent performance (C-index: 0.869, 95% CI: 0.816-0.922; HR: 13.55, 95% CI: 7.409-24.78). Conclusion:Our study identified a novel gene-pair signature for prediction of RFS in CC.

SUBMITTER: Chen PF

PROVIDER: S-EPMC6175542 | biostudies-literature | 2018

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

A novel gene-pair signature for relapse-free survival prediction in colon cancer.

Chen Peng-Fei PF Wang Fan F Zhang Zi-Xiong ZX Nie Jia-Yan JY Liu Lan L Feng Jue-Rong JR Zhou Rui R Wang Hong-Ling HL Liu Jing J Zhao Qiu Q

Cancer management and research 20181003

<h4>Background</h4>Colon cancer (CC) patients with early relapse usually have a poor prognosis. In this study, we aimed to identify a novel signature to improve the prediction of relapse-free survival (RFS) in CC.<h4>Methods</h4>Four microarray datasets were merged into a training set (n=1,045), and one RNA-sequencing dataset was used as a validation set (n=384). In the training set, microarray meta-analysis screened out 596 common RFS-related genes across datasets, which were used to construct ...[more]

PMID: 30323670

Similar Datasets

Project description:Purpose: The benefit of postoperative adjuvant chemotherapy in patients with Dukes´ B colorectal cancer is still uncertain and its routine use is not recommended. The five years survival rate is approximately 75% and identification of the patients at high risk of recurrence would represent an important strategy for the use of adjuvant chemotherapy. In this study we identify new prognostic markers for tumor relapse in Dukes´ B colon cancer patients. Patients and Methods: We retrospectively analyzed gene expression profiles in frozen tumor specimens from 16 patients with Dukes´ B colorectal cancer by using high density oligonucleotide microarrays. Data were normalized and subsequently analyzed with two different statistical procedures. The intensity value associated to each spot is the result of subtracting a gaussian function of the noise from the foreground values {Kooperberg, 2002}. After this background subtraction, base 2 logarithms of all data were calculated and genes with more than two missing values were excluded from the analysis. The remaining missing values were replaced by using the KNN imputation method {Troyanskaya, 2001}. The great number of genes present in the microarray allowed us to consider that the overall fluorescence intensity must be the same for all slides, as the vast majority of genes will not change their expression between the two conditions tested. So we used the quantile normalization method {Bolstad, 2003}, which equalizes not only the average intensity but also the range of intensity values between slides. We decided to use two different statistical procedures in the search of significant differences in gene expression between the relapsed and non-relapsed patients. Only the genes selected by both statistical procedures were considered as differentially expressed between relapsed and non-relapsed patients. The first test was a permutation t-test for comparison of two means {Dudoit, 2003} and the second one a variation of the Fisher test based on the work presented by Iizuka et al. {Iizuka, 2003} in which they searched for the optimal number of genes that could differentiate between two groups of samples. Briefly, we used the same algorithms for the calculation of the Fisher criterion, but instead of looking for the optimal subset of genes able to differentiate between groups, what requires a lot of computing capacity, we tried different numbers of candidate genes and selected a number which yielded very good classification results when evaluated by means of Fast ICA {Lee, 2003} and Hierarchical Clustering {Everitt, 1974}. The procedure consisted of selecting the 30 genes with the highest Fisher criterion value in 6 rounds of iteration, each round leaving one sample of the relapsed group and two of the non-relapsed group out of the calculations (a variant of the “leave one out” method). We selected the genes present in at least 3 of the iterations. The intersection of the groups of genes selected by the two statistical procedures was selected as a prognostic signature for relapse in Dukes´stage B colon cancer. Results: Our results show a group of 48 genes differentially expressed between the relapsed and non-relapsed groups with an associated probability below 0,001 in the t test. Another statistical procedure based on the Fisher criterion resulted in 11 genes able to separate both groups. In order to minimize false positives we only considered a good gene signature the 8 genes selected by both statistical procedures. These genes are ribosomal protein S5, chromodomain helicase DNA binding protein 2, lysosomal ATPase V0 subunit A isoform 1, zinc finger protein 148, brain protein I3 (BRI3), hypothetical protein MGC23401 and one unknown clone. In order to verify the obtained results, the differential expression of the first two genes was confirmed by real time PCR. Keywords: repeat sample

Project description:BackgroundMetastatic colorectal cancer (mCRC) is a highly heterogeneous disease from a clinical, molecular, and immunological perspective. Current predictive models rely primarily in tissue based genetic analysis, which not always correlate with inflammatory response. Here we evaluated the role of a circulating inflammatory signature as a prognostic marker in mCRC.MethodsTwo hundred eleven newly diagnosed patients with mCRC were enrolled in the study. One hundred twenty-one patients had unresectable metastases, whereas ninety patients had potentially resectable liver metastases at presentation. Analysis of miR-21, IL-6, and IL-8 in the plasma of peripheral blood was performed at baseline. Patients with high circulating levels of ≥2 of the three inflammation markers (miR-21, IL-6, and IL-8) were considered to have the "Inflammation phenotype-positive CISIG".ResultsPositive CISIG was found in 39/90 (43%) and 50/121 (45%) patients in the resectable and unresectable cohort, respectively. In the resectable population the median relapse-free survival was 18.4 compared to 31.4 months (p = 0.001 HR 2.09, 95% CI 1.2-3.67) for positive vs. negative CISIG. In contrast, the individual components were not significant. In the same population the median overall survival was 46.2 compared to 66.0 months (p = 0.0003, HR 2.57, 95% CI 1.26-5.27) for positive vs. negative CISIG, but not significant for the individual components. In the unresectable population, the median overall survival was 13.5 compared to 25.0 months (p = 0.0008, HR 2.49, 95% CI 1.46-4.22) for positive vs. negative CISIG. IL-6 was independently prognostic with overall survival of 16.2 compared to 27.0 months (p = 0.004, HR 1.96, 95% CI 1.24-3.11) for high vs. low IL-6, but not the other components. Using a Cox regression model, we demonstrated that CISIG is an independent predictive marker of survival in patients with unresectable disease (HR 1.8, 95% CI 1.2, 2.8, p < 0.01).ConclusionIn two different cohorts, we demonstrated that CISIG is a strong prognostic factor of relapse-free and overall survival of patients with mCRC. Based on these data, analysis of circulating inflammatory signaling can be complimentary to traditional molecular testing.

Dataset Information

A novel gene-pair signature for relapse-free survival prediction in colon cancer.

Publications

A novel gene-pair signature for relapse-free survival prediction in colon cancer.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure