Page 32 - GPD-2-3

Gene & Protein in Disease Signatures construction strategies for TC

cohort. For instance, the internal validation was of TC by constructing the risk assessment models:
performed using transcriptome and clinical data of PTC nomograms [19,27,36,38,56,78] .
patients derived from the Nucleotide Archive database, Nowadays, nomogram can be widely used for combining
and the external validation was employed by inducing diagnosis with multiple indexes and also predict the
11 patients who had undergone total thyroidectomy in recurrence and progression of many carcinomas, including
Beijing Cancer Hospital . In addition, Liu et al. and Lv [19,38,56] [78] [36] [27]
[21]
et al. acquired the patients’ data from the hospitals and PFI , RFS , DFS , and OS for DTC, PTC, or TC
performed further validation in the regulation of each patients. Based on the prognostic signature established
molecule of the signature [29,47] . In addition to statistical by the multivariate regression hazard model and some
validation methods, some studies also used experimental relative clinical parameters, the nomogram could be
validation methods such as immunohistochemistry (IHC) developed. The assessment indexes including the C-index,
and quantitative real-time polymerase chain reaction AUC of ROC, and the calibration plots could evaluate the
(qRT-PCR). As effective external validation methods, performance of the nomogram [19,36,38,78] . For instance, Pan
IHC and qRT-PCR were used to verify the differential et al. screened 5 genes to establish a prognostic signature
expression of the biomarkers during the construction and and then validated its prognostic value for clinical
validation of signatures [25,29,31,33,45,79] . As common methods applications. In addition, a nomogram of predicting
of experimental verification, both of them could further DFS of PTC patients was constructed with a C-index of
verify the expression level of molecules in prognostic tags, 0.797 (95% CI, 0.730–0.864), and AUCs for 1-, 3-, and
[36]
making the signatures more convincing. 5-year DFS were 0.763, 0.777, and 0.755, respectively .
Chen et al. identified a DNA-methylated signature that
3.5. Nomogram based on prognostic signatures for TC has a good performance for RFS of PTC and then based
To demonstrate the prognostic abilities of signatures, some on the risk score and multivariate regression model, a
studies predicted the risk of various outcome variables nomogram utilized in the clinic was constructed . Wang
[78]

Table 5. Nomograms constructed for prognostic signatures of TC patients
Author Signatures Pathological Survival Nomogram validation methods References
type type event AUCs of the ROC Harrell’s P‑value Calibration
curve (95% CI) concordance of K‑M plot
index (95% CI) analysis
Ruchong Pan, mRNA DTC DFS AUCs for 1-, 3-, and 0.797 (95% CI, Yes Ruchong
et al., 2021 5-year DFS: 0.763, 0.730 – 0.864) et al., 2021 [36]
0.777, and 0.755
Zhiwei Chen, mRNA TC OS Yes Chen
et al., 2021 et al., 2021 [79]
Mingqin Ge, mRNA TC OS Yes Ge
et al., 2021 et al., 2021 [27]
Mengwei Wu, mRNA PTC PFI The AUCs for the 1-, 0.790 (95% CI, P<0.0001 Yes Wu
et al., 2020 3-, and 5-year PFI: 0.652 – 0.927) et al., 2020 [38]
0.855 (95% CI, 0.779
– 0.932), 0.799 (95%
CI, 0.722 – 0.877),
and 0.812 (95% CI,
0.718 – 0.907)
Mengwei Wu, mRNA PTC PFI The AUCs for the 1-, 0.7600 (95% CI, P<0.0001 Yes Wu
et al., 2019 3-, and 5-year PFI: 0.6759 – 0.8440) et al., 2019 [19]
0.7480, 0.7550, and
0.7627
Kun Wang, mRNA/ PTC PFI 0.792 (95% CI, Yes Wang
et al., 2020 lncRNA 0.716 – 0.867) et al., 2020 [56]
Hengyu Chen, DNA PTC RFS The AUCs for 1-, 0.796 (95% CI, Yes Chen
et al., 2020 methylation 3-, and 5-year RFS: 0.704 – 0.888) et al., 2020 [78]
0.850, 0.783, and
0.800
Abbreviations: PFI: Progression-free intervals; DFS: Disease-free survival; RFS: Recurrence-free survival; OS: Overall survival

Volume 2 Issue 3 (2023) 16 https://doi.org/10.36922/gpd.1138