ColoGuide_Stage_II-Msig signature reporting file

Generated by ReProMSig version 1.0

Date created: 2022-03-23

  Last update: 2023-07-18



Lead contact: Jianmin Wu

Email address: wujm@bjmu.edu.cn

Organization: Peking University Cancer Hospital & Institute

Reporting file URL: https://omics.bjcancer.org/prognosis/open/ColoGuide_Stage_II-Msig_a1cffaed_report.html

I. Signature summary

Signature name: ColoGuide_Stage_II-Msig

Signature type: Prognostic molecular signature

Signature generation: 

Patient risk stratification: 2 Groups

Signature description: A 10-gene expression signature for stage II colorectal cancer prognosis, developed and evaluated by ReProMSig, using published gene expression datasets and clinicopathological information.


Training & Validation datasets

Class Dataset name Primary site Disease type Molecular dataset Endpoint Pubmed Clinical trial Registry
Training dataset ProLT_stageII Colorectal Adenomas and Adenocarcinomas (96) GSE30378 , GSE24550 (96 in total) DFS (30 events, 96 patients) 22991413
Validation dataset ProLT_stageIII Colorectal Adenomas and Adenocarcinomas (76) GSE30378 , GSE24550 (76 in total) DFS (37 events, 76 patients) 22991413
Validation dataset ProV_stageII Colorectal Adenomas and Adenocarcinomas (108) GSE14333 , GSE17538 (108 in total) DFS (14 events, 108 patients) 19996206 , 19914252
Validation dataset ProV_stageIII Colorectal Adenomas and Adenocarcinomas (110) GSE14333 , GSE17538 (110 in total) DFS (37 events, 110 patients) 19996206 , 19914252


II. Items of methods

Item 4. Source of data

Item 4a. Study design

Training dataset

  • Study design or source of data

Sourced from 2 independent series of altogether 96 patients with stage II CRC.
- 52 samples from GSE30378 were taken from patients treated surgically at different hospitals in the Oslo region, before adjuvant chemotherapy becoming standard treatment for stage III patients;
- 44 samples from GSE24550 were consecutively collected from patients treated by curative resection at a Norwegian hospital (Aker University Hospital, Oslo, Norway).

Validation dataset(s)

  • Study design or source of data

ProLT_stageIII: sourced from 2 independent series of altogether 76 patients with stage III CRC.
- 43 samples from GSE30378 were taken from patients treated surgically at different hospitals in the Oslo region, before adjuvant chemotherapy becoming standard treatment for stage III patients;
- 33 samples from GSE24550 were consecutively collected from patients treated by curative resection at a Norwegian hospital (Aker University Hospital, Oslo, Norway).

ProV_stageII: sourced from 2 independent series of altogether 108 patients with stage II CRC.
- Consecutive patients from GSE14333 (n = 94) were retrieved from the tissue banks of the Royal Melbourne Hospital and the H. Lee Moffitt Cancer Center in the United States.
- Patients from GSE17538 (n = 14) were treated at the Vanderbilt Medical Centre (Nashville, TN, USA).

ProV_stageIII: sourced from 2 independent series of altogether 110 patients with stage III CRC.
- Consecutive patients from GSE14333 (n = 91) were retrieved from the tissue banks of the Royal Melbourne Hospital and the H. Lee Moffitt Cancer Center in the United States.
- Patients from GSE17538 (n = 19) were treated at the Vanderbilt Medical Centre (Nashville, TN, USA).

Item 4b. Key study dates

Training dataset

52 patients from GSE30378 were recruited from 1987 to 1989, and 44 patients from GSE24550 were recruited from 2005 to 2008.

Validation dataset(s)

ProLT_stageIII: 43 patients from GSE30378 were recruited from 1987 to 1989, and 33 patients from GSE24550 were recruited from 2005 to 2008.

ProV_stageII: Not mentioned in this dataset.

ProV_stageIII: Not mentioned in this dataset.


Item 5. Participants

Item 5a. Key study setting

Training dataset

  • Study setting, number and location of centers

- 52 patients were collected from hospitals in the Oslo region.
- 44 patients were consecutively collected at Oslo University Hospital, Aker, Norway.

Validation dataset(s)

  • Study setting, number and location of centers

ProLT_stageIII:
- 43 patients were collected from hospitals in the Oslo region.
- 33 patients were consecutively collected at Oslo University Hospital, Aker, Norway.

ProV_stageII: Patients were collected at H Lee Moffitt Cancer Center, Royal Melbourne Hospital, and Vanderbilt University Medical Center.

ProV_stageIII: Same as the ProV_stageII dataset.

Item 5b. Eligibility criteria for participants

Training dataset

  • Exclusion criteria

Clinical anlaysis

Patients with stage III tumors were excluded.

Statistical anlaysis

No patients were excluded by ReProMSig.

  • Inclusion criteria

Clinical anlaysis

Patients with stage II tumors were selected.

Statistical anlaysis

All patients were included in ReProMSig.

Validation dataset(s)

  • Exclusion criteria

Clinical anlaysis

ProLT_stageIII: Patients with stage II tumors were excluded.

ProV_stageII: Patients with stage III tumors were excluded. Individuals from GSE14333 series who had received preoperative chemotherapy and/or radiotherapy or for whom tumor-derived total RNA was inadequate for microarray analysis (RNA integrity number [RIN] < 6) were excluded.

ProV_stageIII: Patients with stage II tumors were excluded. Individuals from GSE14333 series who had received preoperative chemotherapy and/or radiotherapy or for whom tumor-derived total RNA was inadequate for microarray analysis (RNA integrity number [RIN] < 6) were excluded.

Statistical anlaysis

No patients were excluded by ReProMSig.

  • Inclusion criteria

Clinical anlaysis

ProLT_stageIII: Stage III patients were selected.

ProV_stageII: Nonoverlapping stage II patients between GSE14333 and GSE17538 series were included.

ProV_stageIII: Nonoverlapping stage III patients between GSE14333 and GSE17538 series were included.

Statistical anlaysis

All patients were included in ReProMSig.

  • Is it similar to the eligibility criteria used in training dataset

    • Similar as aforementioned.

    Item 5c. Received treatments

    Training dataset

    • Treatment details

    - Patients from GSE30378 series had not received adjuvant chemotherapy, which was introduced as standard treatment for patients with stage III CRC aged <75 years in Norway in 1997.

    - Patients from GSE24550 had not received preoperative radiotherapy and adjuvant chemotherapy was given in accordance with Norwegian guidelines to patients with stage III colon cancer aged <75 years, or to patients with stage II disease in whom <8 lymph nodes were examined, or to patients with preoperative or intraoperative tumour perforation. Furthermore, adjuvant chemotherapy is generally only given to physically fit patients, but not to patients with rectal cancer. All underwent curative resection and no bowel perforation was reported.

    Validation dataset(s)

    • Treatment details

    ProLT_stageIII: Same as the training dataset.

    ProV_stageII:
    - Patients from GSE14333 series had not received preoperative chemotherapy and/or radiotherapy. Nine patients had received postoperative concurrent chemoradiotherapy (50.4 Gy in 28 fractions with concurrent 5-fluorouracil), 13 patients had only received adjuvant chemotherapy.
    - 14 patients from GSE17538 had no available information about received treatments.

    ProV_stageIII:
    - Patients from GSE14333 series had not received preoperative chemotherapy and/or radiotherapy. 12 patients had received postoperative concurrent chemoradiotherapy (50.4 Gy in 28 fractions with concurrent 5-fluorouracil), one patient had only received adjuvant radiotherapy, 51 patients had only received adjuvant chemotherapy (either single agent 5-fluouracil/capecitabine or 5-fluouracil and oxaliplatin).
    - 19 patients from GSE17538 series had no available information about received treatments.


    Item 6. Outcome

    Item 6a. Clinical endpoints being analyzed

    • Outcome

    Relapse or death from colorectal cancer was regarded as events, and patients with no events were censored. Specially, 10 years follow-up was done for the GSE30378 dataset, and 5 years follow-up was done for the GSE24550 dataset.

    • How and when assessed

    Not mentioned in this study.

    Item 6b. Actions for blind assessment of the outcome

    Not mentioned in this study.


    Item 7. Predictors

    Item 7a. Predictors used in developing the multivariable prediction model

    Genome-wide expression at the exon-level for colorectal cancer tissue biopsies was analyzed using the Affymetrix Human Exon 1.0 ST platform.

    Item 7b. Actions for blind assessment of predictors

    Not mentioned in this study.


    Item 8. Sample Size

    • How the study size arrived at

    Not mentioned in this study.


    Item 9. Missing Data

    Training dataset

    • Missing value handling

    Assuming occurrence of missing values in gene expression profiles were random events, KNN method (impute.knn function in R package impute version 1.60.0) was applied to impute missing values by ReProMSig.

    Validation dataset(s)

    • Missing value handling

    Assuming occurrence of missing values in gene expression profiles were random events, KNN method (impute.knn function in R package impute version 1.60.0) was applied to impute missing values by ReProMSig.


    Item 10. Statistical Analysis Methods

    Item 10a. Predictors handling method

    • Molecular profiles

    Expression values were converted to log2 transformed. Using training dataset gene expression values as the reference, ComBat function in R package "sva" (version 3.34.0) was applied to reduce the likelihood of batch effects from nonbiological technical biases for each validation dataset profiles.

    • Implausible observations

    Extreme values (e.g., outliers) in expression profiles were regarded as NA, and imputed by KNN method, using impute.knn function in R package "impute " (version 1.60.0).

    Item 10b. Model building procedures

    • Predictor selection before modeling

    The following genes (n = 738) with expression variances higher than 0.2 and P-values from univariate Cox proportional hazards analyses below 0.1 were used for predictor selection and modelling. P-values (Wald test of predictive potential) were calculated in R 2.11.1 using the Bioconductor package Weighted Gene P-values (Wald test of predictive potential) were calculated in R 2.11.1 using the Bioconductor package Weighted Gene Co-expression Network Analysis (WGCNA).
    WNT11 DNASE1 AVL9 STAU1 DKK1 RPL10 ANXA9 CMTM7 DNTTIP1 RHCG CHN2 PPDPF MOGAT3 KCNJ3 NPAS2 SEMA3A SLC15A1 PLOD3 GNG4 B3GNT3 REEP1 RMND5B BCL2L1 MYRIP SLC22A3 SLC5A6 ZDHHC4 HAS2 NTS CR2 GALNT10 TFCP2L1 SQSTM1 PLCG1 TMEM8A HMGN2 DYNLRB1 DDX27 PLCG2 SLCO4A1 KIF3B GPRC5A PPY TULP3 PXMP4 TP53RK PRKCSH HKDC1 SOX6 ZNF470 PLEKHB1 GPR155 FAM50B TM9SF4 PYGB CCL19 HM13 NME4 PLAG1 C10orf11 FGFR4 SLC6A6 SLC1A5 PHGDH GLS MAPRE3 SLC35C2 SERINC5 SH3BP4 GLIPR2 ALG3 DAPK1 RTKN KRTCAP2 EPB41L1 TGOLN2 CRIP1 ZNF426 CDHR2 SLC14A1 NUDT4 AZGP1 CLCN2 SLC17A9 RHPN2 ARFGEF2 MT1M SLC2A8 C15orf48 GREM1 NKD1 GPR39 SRC PTK6 PLEKHG6 SLC18A3 NGEF ZNF217 TNFAIP2 ANKH ADRM1 ULBP1 OR1F1 BAIAP2L1 SMAP1 SLAIN1 GDI1 ATP1B3 ID1 TRIP10 RASA1 LY75 FST C7orf13 TGFBI BAG2 ZNF442 SGK2 SPRR3 SLC41A3 PVRIG CD52 SERPIND1 AREG RBM39 SMCR5 KIAA1191 EREG PCDHB9 LRP5 SNRPF PIGT NDRG3 TJP3 YTHDF1 STC2 IFRD1 ZNF574 SFRP1 LAPTM4B ZNF260 EGR3 SH3TC2 FADS2 CD69 STAMBP ZFP64 PARK7 CD200 CYP24A1 CPNE1 APOA2 APBB1IP WWC2 MAML2 TMEM41A SLC3A2 LAMB3 SLC45A4 TESC SLC7A11 ITM2A TBC1D9B RAMP3 ATP6V0D2 CLDN6 PARD6B NNT GDF15 ZNF581 ARL8A ZNF334 NT5DC3 SPATA2 PRKG2 SESN1 HOXD3 APOL3 OR10A4 TRPC4AP ZFAND2A TLR10 PARP15 NALCN NCK2 PHF3 GSK3B PMEPA1 NEBL SOX14 OR3A3 NCOA3 PHLDA1 FBXW11 ENTPD6 RNF125 NDC80 SLC20A2 TGIF2 ZSWIM3 BCL2 LEMD1 WFDC10A CORO2A GFRA3 BIK MANSC1 FAM127B RBM22 CEACAM1 EFNA1 SMURF1 EXOSC5 SMG5 RALGAPB DOCK8 MED20 EPDR1 ERGIC3 IGF2 FMO4 KIAA1324 PLIN3 SPINT3 LRRC8D ELMO2 ZDHHC9 CREB3L1 UBB LDLRAD3 PDPK1 HOXA7 CXCL13 DPY19L2P2 CTNNBL1 ZNF473 TMC7 MSX2 MDH2 ITGAX GALNT2 ELF5 SHISA5 NSUN5 PLEK F11R PLAGL2 ZKSCAN1 ZNF669 MC1R PRKCB ITGA6 CLTB CNPY3 OSBPL2 NFYA CHCHD2 RAB22A EXOC2 SCAMP5 ZNF570 SPATS2L CTNNA2 SULT2A1 IER3 PI3 NPHP1 LOC554207 ZSCAN20 S100A9 TPD52L2 VPREB3 HNF4A SH3RF1 SLC2A12 TBCK GPR35 STK39 PLAT PCDH1 NFS1 C20orf24 HLA-DRA RPL17 KCNN4 ATIC TMEM183A SERINC3 BTF3 FAM126B CDK5RAP1 CCT6A GYLTL1B CD3E IL12RB2 RNF43 PTTG3P CXCL9 TOB1 SNTB1 CLDN1 EIF4B NOV ZDHHC23 TSPYL4 NR1I2 SELL RUFY1 LNX2 EPS8L3 EPB41L5 PPP1R15A HDAC11 SOSTDC1 C15orf37 TBCB DGAT2 EIF4H PLXNC1 CLCN4 JUP LIN7A UCHL1 RPE H2BFXP EPN1 ZNF788 ZNF713 WDR55 PPM1B PDZD11 EDNRA ZNF354C AGT MEF2D YIPF4 CYP4F11 MAPK14 CEACAM5 ZNF3 TTC37 GLB1 AGR3 PTGDS CLCN3 PRPF6 KRTAP7-1 KCNK17 VPS72 XAB2 NEU1 DDR1 WDR35 PDE5A TTYH3 VNN2 CACNA1D ITGB5 CCL18 PLIN2 ABCC5 NDFIP1 SFT2D3 GATS SF3B4 BCL2A1 DNAJC5 XRCC5 TPX2 TMEM63A ZNF514 PRR15 STX11 HERPUD2 SRPR ROCK2 PAMR1 EIF2AK1 NFAT5 TM4SF1 MACC1 SCFD2 PCSK1 SORT1 NUP43 KCMF1 LARP1 CD48 MTMR12 DYM CTTNBP2 TNFRSF19 SLC9A8 MUC13 CTNNAL1 STT3B MBLAC2 SDC4 SMG7 CD37 SKIL VEGFA UBL4A EDAR DNAJB1 NCF4 TMEM176A PHF14 GCLC C14orf2 PDZK1 ATP10B CLDN4 GUCY1A3 APLN CIITA CDC42BPA H1F0 NCOA5 ABCF1 RBM12B LIPG APOBEC3F POFUT1 CYB5B TUBA4A GSDMB MET SLC38A2 MAPRE1 PBX3 AARS PTGDR TMEM144 RNF128 GSTO1 ZNF302 BCS1L ELMO1 FAM84B COBLL1 TIGD3 G0S2 YBX1 CHML SYNPR B4GALT5 SMOX GGT7 SPRR1A OXGR1 PPAP2C PCYT1A ATP6AP1 MPZL3 UBE2Q1 PTPN3 ASL SLC25A12 EPB41L3 PGC SUN3 MCTP1 MFSD6 MAP7 C17orf78 NR1D2 FABP4 SPINT2 RQCD1 STXBP6 SDC1 MYO10 DHX35 ZNFX1 DAPK2 RPN2 AP3D1 PCDHB12 DDO ITCH GMPS TBXAS1 FAIM3 MBNL2 PGBD4 P2RY10 GNLY WDFY1 SLC44A5 HOXA2 KLHL24 ACOX1 CXCL10 OTC FAM135A BLCAP OGFR TRAP1 PRKAB1 GNPDA1 PKN1 LRMP TIGD1 TMEM14C RPA2 PRCC PNLIPRP2 RPL23 TMED4 PUM2 TAF1A KTI12 PCIF1 LY6G6D ARHGAP15 ZSWIM1 BCAP31 ECM1 PHLDA3 GRINA IFNA10 BLNK PNN ACTN4 NRBP1 PHYH AVPI1 C9orf163 RAB27B TMEM56 COPS7B SPI1 XPR1 GMFG AGPAT6 TMEM185B GGT6 SLC26A3 TMC4 ZNF827 TRAF3IP3 CXXC5 SEZ6L2 FAM134A ENOPH1 DPF2 APOE EAF1 MOCS3 POGK XIAP CYTH4 STRN SLC2A1 UPF3B ADAM19 SPRR1B SCN9A ARF1 CYP2F1 TMEM176B LYPLA2 HCLS1 YWHAB SLC20A1 GALNT4 GTPBP2 NR4A3 LPGAT1 VIP PTPRK CD79A ABCC2 RHBDD1 TTPAL NCOA6 PUS10 EIF6 ESM1 DOCK11 BTF3L4 CYP3A7 ZNF597 HBZ ASNS ADH1B PLA2G2D PLCL2 HIST1H4L PCDH18 SARS TAAR5 STK40 RNF114 FITM2 TOMM34 CHD6 SLC35E1 HK2 TTC22 ELOVL5 RPRD1A BCL2L14 SULF2 SLC4A2 IGLV6-57 CCDC80 PIK3CB UBE2N GPD2 MAT2A PRH1 FA2H CHRDL1 MAPK9 ETV4 TRIM13 B4GALT3 CLPTM1 PECR ZC3H15 MYBL2 SPRY4 SLC30A1 RNF165 ZMYND8 VPS13A TMEM53 ALKBH6 DBNDD2 MGAT4B KHDRBS3 GLRX ATP13A3 AMACR DPP7 SSFA2 GRPR APOL1 NMNAT3 SLFN5 IFNK IFNA17 GPR64 C3 USP4 TMEM182 RPP21 DHTKD1 GBP5 CMTM8 PVR AQP1 LAMA3 UST CSTF1 DAGLA RPS29 RAB7A CD3D KBTBD7 LSP1 OLIG3 CHMP1A CAPN6 TCFL5 GPR56 LSR OSM CDH1 CCDC69 MAPK13 TECR RPL11 HHEX POLR3G PLS1 RNF139 WDR12 TBC1D7 PIAS2 C10orf32 STRA13 RAP1B JAM2 CEACAM8 USP53 DEFB129 THYN1 TWISTNB IQCB1 STEAP4 PSPH RALY SLC16A10 COL4A5 LNX1 PARM1 MINA ASXL1 TREH ABHD12 SLC25A26 HARS VPS41 ZNF234 TP53BP2 MTMR11 CIAO1 FNBP1L TBC1D16 HDLBP MMP8 TRIM2 CBFA2T2 RGS1 GPR143 RPS16 TCL1A PPME1 PSAT1 PHF20 FAT1 TGFBRAP1 RSL1D1 ADRBK1 SCEL SYNGAP1 EPHA2 TMC5 NCBP2 PTP4A1 HOXA11 SYNJ2 GRIN2B EEF2 CEP135

    A bootstrapping procedure was applied to select molecular predictors associated with outcome from the list of candidate predictors shown above. For each bootstrap resample (samples were drawn with replacement keeping the same size as the original training dataset), supervised principal components approach (SPCA, R package "SPCA") was applied. The entire process was repeated 200 times, and predictors with frequency of more than 45% occurrence were taken as potential signature predictors for prognosis.

    • Multivariable prediction model building

    Univariate Cox regression model (R package "survival") was fitted to the expression data (training dataset) of the selected molecular predictors, to identify robust molecular signatures (using 0.05 as the significance level) associated with DFS, which were used to generate the multivariable prediction signature (R package "survival").

    • Collinearity assessment

    The correlations between each pair of selected molecular predictors were estimated using Pearson correlation analysis, to evaluate the presence of collinearity. However, we did not exclude any predictors even if potential collinearity found.

    Item 10c. Prediction method for the validation dataset(s)

    • Signature model

    Individual signature score was calculated by a weighted sum of the predictors in the generated multivariate Cox regression model, in which weights are the corresponding regression coefficients (i.e., log hazard ratios). The signature score for a patient is the log relative hazard compared with a hypothetical patient whose signature score is zero.

    • Nomogram

    Nomogram was not constructed as no clinicopathological variable was selected in the development form.

    • Online research tool

    An online research tool for single patient prediction of risk score (i.e., the exponential of signature score) and DFS probabilities at 12, 36, 60-months using the molecular signature is available from https://omics.bjcancer.org/prognosis/.

    Item 10d. Model performance assessment

    • Independence test

    Univariate and multivariate Cox regression analyses were performed to test whether individual prognosis factor is an independent factor in predicting patients DFS. Both the molecular signature prediction (risk groups) and clinicopathological variable (MSI_status) were included in the independence test.

    • Interaction test

    We did not examine interaction terms but relied on the main effects of the selected.

    • Kaplan-Meier analysis

    Discrimination is visually inspected from the spread of Kaplan-Meier curves for each predicted risk group (high-risk, and low-risk), in the training and validation datasets respectively. Differences in the probability of DFS between risk groups (high risk vs low risk) were tested by the two-sided log-rank test. In addition, Kaplan-Meier plots also presented the total number of patients, the number of events (outcome) for each risk group.

    • Time dependent receiver operating characteristic (ROC) analysis

    12, 36, 60-months time-dependent ROC analysis was performed to examine the prognostic accuracy of the developed model in the training dataset. ROC curves for molecular signature were plotted. An area under the ROC curve (AUC) of 0.5 indicates no discrimination, whereas an AUC of 1.0 indicates perfect discrimination.

    • Prediction error (PE) curve analysis

    PE curve analysis was applied in the training dataset to examine the prognosis prediction error rate, and ten-fold cross-validation cumulative prediction error was computed using Kaplan-Meier estimation as reference. Models with smaller area under the curve indicates a relatively lower error rate. PE curve for molecular signature was plotted.

    • Calibration plots

    The calibration plots at 12, 36, 60-months were used to assess the consistency between the actual and predicted DFS probabilities from the molecular signature in training dataset.

    • DFS

    DFS with 95% confidence intervals at 12, 36, 60-months in the training and validation datasets respectively, were calculated by Kaplan-Meier method for patients in each risk group.

    • Association analysis

    To assess whether the developed prediction model is correlated with clinicopathological variables, two-sided Fisher's exact test (for categorical variables) and two-sided t test (for continuous variables) were applied to measure association or difference between predicted risk groups.

    Item 10e. Model updating arising from the validation


    Item 11. Risk Groups

    Patients were stratified into two risk groups on the basis of signature score distribution in each dataset: low-risk (signature score < 0.858) and high-risk (signature score >= 0.858).


    Item 12. Development vs. Validation

    Differences in development and validation

  • Setting

    • The details are described at Item 5a.

  • Eligibility criteria

    • The details are described at Item 5b.

  • Outcome

    • The details are described at Item 6.

  • Predictors

    • The mRNA expression profiles of predictors in training dataset and ProLT_stageIII dataset were generated using Affymetrix Human Exon 1.0 ST platform, while those in ProV_stageII and ProV_stageIII dataset were determined using Affymetrix Human Genome U133Plus 2.0 arrays.



    III. Items of results

    Item 13. Participants

    Item 13a. Study participants

    Training dataset

    In the training dataset, 96 out of 96 (100%) had follow-up of DFS with a median follow-up (by reverse Kaplan-Meier method) of 59.64 months (range: 5.04-120).

    Validation dataset(s)

    ProLT_stageIII: 76 out of 76 (100%) had follow-up of DFS with a median follow-up (by reverse Kaplan-Meier method) of 56.04 months (range: 2.04-120).

    ProV_stageII: 108 out of 108 (100%) had follow-up of DFS with a median follow-up (by reverse Kaplan-Meier method) of 44.8 months (range: 0.43-118.58).

    ProV_stageIII: 110 out of 110 (100%) had follow-up of DFS with a median follow-up (by reverse Kaplan-Meier method) of 52.86 months (range: 0.49-112.33).

    Item 13b/c. Comparison of participant characteristics between the training and validation datasets

    Clinical and pathological characteristics

    Key clinical and pathological characteristics of participants were included. P denotes the P value of the characteristics distribution comparison between the training dataset and each validation dataset. Categorical variables were compared using fisher'exact test, and continuous variables using t-test.


    Expression profile distribution

    The density plots present the distributions of the log2-transformed expression level for the training dataset ProLT_stageII, and the validation datasets ProLT_stageIII, ProV_stageII, ProV_stageIII.


    Item 14. Model Development

    Unadjusted association between each candidate predictor and outcome in the training dataset

    Unadjusted association between top 100 candidate predictors and DFS.


    TRIPOD relevance: This table reports the information needed suggested by Item 14a "Number of participants and outcome events involved in model development" and Item 14b "Unadjusted association between each candidate predictor and outcome".


    Item 15. Model Specification

    Item 15a. Details of the prediction model

    Signature genes and the corresponding coefficients

    10 variables comprising the developed signature and their corresponding coefficients were calculated by the Cox regression model.
    The signature score of a patient can be calculated using the following formula:
    Signature score = -0.047*CCL19 + 0.448*CHN2 - 0.509*CR2 + 0.632*DKK1 + 0.12*GNG4 - 0.365*GREM1 + 0.213*HAS2 - 0.376*MT1M + 0.058*SEMA3A + 0.417*WNT11.
    A higher signature score indicates a relative poorer prognosis.


    Pearson correlation plot for pairwise expression comparison among signature genes

    The correlation plot shows the Pearson correlation coefficients of expression profiles between each pair of signature predictors in the training dataset. From the plot, the correlation pattern of signature predictors can be visually checked if their expression levels are independent. The correlation coefficient ranges from -1 to 1. A smaller absolute value implies that a lower linear dependency, indicating this pair of signature predictors maybe is independent.


    Item 15b. Application of the prediction model

    Nomogram was not constructed as no clinicopathological variable was selected in the development form.


    Item 16. Model Performance

    Cox regression analysis of signature and clinicopathological variables (training dataset)

    Univariate hazard ratio and multivariate hazard ratio of each variable with a 95% confidence interval provided in parentheses, indicate multiplicative effects on the hazard. P denotes the statistical significances in the hazard ratio between a test group relative to a reference group, which were tested by the wald test. For multivariate Cox analysis, the relationship between the variable of interest and the outcome was evaluated after adjusting for potential confounding variables that may be related to the outcome.


    Receiver operating characteristic (ROC) analysis (training dataset)

    Time-dependent receiver operating characteristic (ROC) curves show the sensitivity and specificity of different variables in prognosis prediction. It could help to quantify and compare the discrimination ability of the signature and clinicopathological prognosis factors using the 12, 36, 60-months time-dependent ROC analysis in the training dataset. The area under the curve (AUC) ranges from 0.5 (no discrimination) to a theoretical maximum of 1, which were texted in the legend for each prognosis model. A prognosis model with larger AUC indicates a better performance.


    Calibration of the molecular signature (training dataset)

    Calibration plots at 12, 36, 60-months were generated to explore the performance of the signature. Signature-predicted probabilities and observed outcome were plotted on the X-axis and Y-axis, respectively. A calibration plot along the 45-degree line indicates perfect consistency between the actual and signature-predicted prognosis. The vertical bars represent 95% CIs.


    Prediction error (PE) curve analysis (training dataset)

    Prediction error (PE) curves were used to evaluate the performance of prediction models in the training dataset, and ten‐fold cross‐validation cumulative prediction error were computed using Kaplan-Meier estimation as reference. Integrated Brier score (IBScore) was defined as the area under a prediction error curve, which were texted in the legend for each prognosis model. A prognosis model with smaller IBScore indicates a better performance.


    Association analysis of signature-predicted risk groups with clinicopathological variables

    Association analysis was employed to investigate whether the signature is correlated with clinicopathological variables and other molecular features in both training and validation dataset. Samples with valid values (such as non-missing, non-unknown) were compared in the analysis. P denotes the P values comparing the associations between the clinicopathological/molecular variables and the risk groups in each dataset. Fisher'exact test was used for association analysis.
    ProV_stageII, ProV_stageIII datasets were not shown due to lack of MSI_status variable.


    Kaplan-Meier analysis of signature-predicted risk groups

    Training dataset

    Validation dataset 1

    Validation dataset 2

    Validation dataset 3

    Kaplan-Meier survival curves for DFS in training and validation datasets, stratified by the prediction model (high-risk and low risk). The performance of the prognostic signature was evaluated by the two‐sided log‐rank test in both training dataset and validation datasets. P < 0.05 was considered as statistically significant, and 95% confidence intervals are presented in brackets. HR, hazard ratio.


    12/36/60 months DFS for different signature-predicted risk groups

    12/36/60 months DFS of patients in different risk groups for each dataset were predicted by the prediction model.