br Colon Tumor Side by Gene Expression
Colon Tumor Side by Gene Expression
Figure 2 The Proportions of Samples in Each of the Categories Defined by Anatomic and Genomic Location for Multiple Clinico-pathologic Features Were Plotted. Genomic Location Is a Significant Predictor of Within the Anatomic Left-side Tumors for MSI Status (P [ 1.4 3 10L13) (A), CIMP Status (P [ 1.1 3 10L 14) (B), and BRAF EPZ031686 Status (P [ 6.7 3 10L19) (C). The Differences in Distribution With Respect to Genomic Location and Anatomic Location Are Not Statistically Significant for KRAS Mutation Status (D), Patient Gender (E), and Tumor Stage (F)
Abbreviations: aL-gL ¼ anatomic left e genomic left; aL-gR ¼ anatomic left e genomic right; aR-gL ¼ anatomic right e genomic left; aR-gR ¼ anatomic right e genomic right; CIMP ¼ CpG island methylator phenotype; gNC ¼ genomic noncensensus; MSI ¼ microsatellite instable; MSS ¼ microsatellite stable.
they predict the genomic left samples using a linear model in the Cohort A training set. Candidate scores were computed (see Methods) using the highest ranked 2, 5, 7, 10, or 15 risk scores and assessed as predictors of the genomic location using the receiver operating characteristic method. We evaluated each score S as a predictor of the genomic left samples, and 1-S as a predictor of the genomic right samples. The highest area under the curve (AUC) values for both left (AUC ¼ 0.95) and right (AUC ¼ 0.94) were
Emily Cannon, Steven Buechler
Figure 3 The Proportions of Samples in Each of the Categories Defined by Anatomic and Genomic Location for the Consensus Molecular Subtypes Were Plotted. The Proportion of CMS4 Samples in gNC (40%) Is Significantly (P [ 1.3 3 10L7) Greater Than in Other Tumors (18%). Genomic Location Is Also Superior to Anatomic Location in Predicting CMS1 (Right-side Dominant) and CMS2 (Left-side Dominant)
Abbreviations: aL-gL ¼ anatomic left e genomic left; aL-gR ¼ anatomic left e genomic right; aR-gL ¼ anatomic right e genomic left; aR-gR ¼ anatomic right e genomic right; CMS ¼ consensus molecular subtype; gNC ¼ genomic noncensensus.
obtained with the 7-probe score (Table 4), hereafter called the genomic location score. In the Cohort A validation set, the genomic location score predicted genomic location with AUC ¼ 0.90 for genomic left-side tumors (Figure 4A) and AUC ¼ 0.91 for genomic right-side tumors (Figure 4B).
Owing to the high degree of overlap between genomic location and anatomic location, the genomic location score is also predictive of anatomic location (AUC ¼ 0.90, in the Cohort A validation set). The ability of genomic location score to predict genomic location goes beyond prediction of anatomic location as follows. In the
Cohort A validation set, the genomic location score predicted genomic left-side tumors within the anatomic right-side tumors (Figure 4C) (AUC ¼ 0.86) and genomic right-side tumors within the anatomic left-side tumors (Figure 4D) (AUC ¼ 0.89). Score values for each sample in Cohort A are found in Supplemental Table 1 (in the online version).
To show that the utility of the genomic location score is not limited to Cohort A, it was computed for samples in Cohort B using gene expression values for the 7-panel genes (Table 4) assessed by RNA-seq (see Methods). The genomic location score predicted
Colon Tumor Side by Gene Expression
Table 4 Genes Used to Compute Genomic Location Score
Feature ENTREZID SYMBOL Elevation
anatomic location in Cohort B with AUC ¼ 0.92. Thus, the relationship between genomic location score and anatomic location was comparable in Cohort B and Cohort A validation set.
Tumor location is increasingly a consideration in planning the treatment of colon cancer.18 Because anti-tumor drugs act at the molecular level, there is need for greater knowledge of the molecular differences between left- and right-side colon cancer. In this study, we analyzed the degree to which gene expression determines the side of a tumor. Using a methodology that reduces bias owing to sta-tistical methodology, we showed that gene expression predicted tumor location with 81% accuracy; however, 6% of tumors were
Figure 4 Receiver Operator Characteristic Curves Were Plotted in the Cohort A Validation Set for the 7-Probe Genomic Location Score as a Predictor of Genomic Left-side Tumors (A), Genomic Right-side Tumors (B), Genomic Left-side Tumors Within the Anatomic Right-side Tumors (C), and Genomic Right-side Tumors Within the Anatomic Left-side Tumors (D)
1 – specificity 1 – specificity D