Hindawi

Journal of Oncology
Volume 2023, Article ID 2733232, 16 pages
https://doi.org/10.1155/2023/2733232

Research Article
Construction of a Prognostic Model for Predicting Colorectal
Cancer Prognosis and Response to Immunotherapy Based on
Cuproptosis-Associated lncRNAs

Yi Yang ,1,2 Xiaoli Wang ,3 Jin Lu ,4 Zhiyong Dong ,1 Ruixiang Hu ,1

Wenhui Chen ,1 Songhao Hu ,1 Guanhua Lu ,1 Biao Huang ,1 Shiliang Dong ,1
Lu Wang ,5 and Cunchuan Wang 1,2
1
Department of Bariatric Surgery, Te First Afliated Hospital of Jinan University, Guangzhou, Guangdong, China
2
Jinan University Institute of Obesity and Metabolic Disorders, Guangzhou, Guangdong, China
3
West General Department, Te First Afliated Hospital of Jinan University, Guangzhou, China
4
Key Laboratory of Computational Medicine and Intelligent Health of Anhui Higher Education Institutes,
Bengbu Medical College, Bengbu, Anhui, China
5
Institute of Precision Cancer Medicine and Pathology, Jinan University Medical College, Guangzhou, Guangdong, China

Correspondence should be addressed to Lu Wang; luwang2022@jnu.edu.cn and Cunchuan Wang; twcc@jnu.edu.cn

Received 26 July 2022; Revised 29 August 2022; Accepted 30 August 2022; Published 15 March 2023

Academic Editor: Bin Liu

Copyright © 2023 Yi Yang et al. Tis is an open access article distributed under the Creative Commons Attribution License, which
permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Colorectal cancer (CRC) is a common and highly lethal gastrointestinal malignancy. Immunotherapy has shown positive efcacy
in the treatment of CRC; however, only a minority of patients beneft from immunotherapy. Te aim of this study is to construct
a cuproptosis-related lncRNA (CRLs) risk score model to predict the prognosis and immune infltration of CRC patients. Firstly,
we synthetically analyzed 19 cuproptosis-related genes (CRGs) from CRC samples derived from the TCGA and obtained 33 CRLs
that were signifcantly associated with prognosis. Next, we defned three cuproptosis modifcation patterns via consensus
clustering analysis (C1, C2, and C3). Further analysis showed that there were signifcant diferences in the abundance of B cells,
NK cells, fbroblasts, monocytes, CD8+ cells, bone marrow dendritic cells, and cytotoxic lymphocytes in diferent clusters. In
addition, the LASSO regression screened out 6 individual CRLs (AC009315.1, PLS3-AS1, ZEB1-AS1, AC007608.3, AC010789.2,
and AC010207.1) closely related to the prognosis of CRC. We found that the low-risk group had better survival prognoses in
patients. Furthermore, the high-risk group had lower immune scores and exhibited lower CD8+ T cell infltration. Moreover, the
low-risk group had lower immune exclusion, immune dysfunction and TIDE scores than the high-risk group. Interestingly, the
lncRNAs in our risk model were positively associated with most immune checkpoints. CD274 (PD-L1), CTLA4, and HAVCR2
(TIM3) were positively correlated with risk scores. Moreover, MSI-H patients had lower risk scores than MSI-L patients, and IPS
scores were signifcantly higher in the low CRLs score group. In conclusion, we constructed a novel risk score model with6
lncRNAs related to cuproptosis, which may be a potential biomarker for evaluating the prognosis and immune treatment for CRC.

1. Introduction However, the prognosis of patients with advanced CRC

In the past, surgery and chemotherapy were the main
treatment methods for colorectal cancer (CRC) [1]. In recent
years, with the advent of immunotherapy, targeted therapy
and other treatment strategies, the prognosis of colorectal
cancer patients has been signifcantly improved [2].
remains poor, largely due to the lack of highly specifc
prognostic biomarkers [3]. So far, the TNM staging system is
the most commonly used prognostic indicator in clinical
practice, but its overall specifcity is insufcient [4].
Terefore, it is crucial to explore more sensitive and specifc
markers for the prognosis of CRC.
2 Journal of Oncology
Cuproptosis is a newly discovered form of programmed
cell death, which is diferent from known programmed cell
death such as apoptosis, ferroptosis, pyroptosis, and nec-
roptosis; relies on intracellular overload of copper ions to
cause cellular death. Excessive respiration produces cyto-
toxicity and eventually induces cell death [5]. Recent studies
have shown that cuproptosis regulation is involved in the
development and response to therapy of multiple tumor
types [6–8]. Numerous proteins, such as CDKN2A, FDX1,
DLD, DLAT, LIAS, GLS, LIPT1, MTF1, PDHA1, and PDHB,
have been identifed that afect tumor cell proliferation and
migration and are associated with cuproptosis [9]. Tere-
fore, revealing the occurrence and development mechanism
of cuproptosis may provide positive help for the treatment
of CRC.
In recent years, immunotherapy has emerged as
a promising alternative therapy for CRC patients. However,
due to tumor heterogeneity, only a minority of patients
beneft from immunotherapy [10]. Given the evidence
suggests that intratumoral infltrating leukocytes are closely
associated with the efciency of immune responses, in-
cluding in CRC [11]. Terefore, the discovery and identi-
fcation of novel immune-related gene targets are crucial to
accurately predict the immune response of CRC.
LncRNAs are RNAs containing more than 200 nucle-
otides that cannot be translated into proteins [12]. LncRNAs
play an important role in the occurrence and progression of
various solid cancers, including CRC [13–15]. More in-
terestingly, a series of prognostic models constructed based
on public databases and by analyzing the expression of
lncRNAs showed excellent predictive ability [16, 17]. A
prognostic model based on a collection of various regulatory
functions in tumors may be a positive direction for the
exploration of prognostic markers in the future. However,
there has been no report on the construction of a prognostic
model based on cuproptosis-related lncRNAs (CRLs).
In this study, we obtained RNA-sequencing (RNA-seq)
data from the TCGA database and identifed 6CRLs sig-
nifcantly associated with prognostic and then developed
a prognostic model. In addition, we verifed the CRLs model
with training and validation cohort and explored its un-
derlying mechanisms through enrichment analysis. Finally,
we assessed the relationship between risk scores and im-
mune cell infltration, drug sensitivity, and immunotherapy
efcacy. Our fndings will help predict the prognosis of
colorectal cancer patients and provide references for clinical
immunotherapy.
2. Materials and Method
2.1. Data Collection and Correlation Analysis. Te Cancer
Genome Atlas (TCGA) database was used to retrieve the
RNA transcriptome dataset and the associated CRC clinical
data. Genes were divided into protein-coding genes and
lncRNA genes based on information from the annotated
human genome. Additionally, the levels of 19 cuproptosis-
related genes (CRGs) expression were evaluated. To evaluate
the relationship between lncRNAs and CRGs, we employed
Pearson correlation coefcients. CRLs were those with an
absolute correlation coefcient of >0.4 and a p value less
than 0.001. After that, patients were split into the training
group and the validation group. Te data that were retrieved
were then used for bioinformatics analysis.
2.2. Construction of Risk Model. To fnd lncRNA predictive
characteristics connected to cuproptosis in the training data
set, univariate Cox regression analysis and minimal absolute
shrinkage and selection operator (lasso) penalized Cox re-
gression analysis were utilized. Each CRC patient’s risk score
was determined using the following formula: Risk score is
equal to Expi ∗ i, where Expi and bi are the expression and
coefcient of each lncRNA, respectively.
2.3. Estimation of Tumor-Microenvironment Cell Infltration.
In this study, we applied the method of Cell type Identif-
cation By Estimating Relative Subsets Of RNA Transcripts
(CIBERSORT) to quantify 22 types of immune cells in the
tumor and normal tissue [18]. We also applied the micro-
environmental cell population counter (MCPcecther)
method using the R package “MCPcether” to quantify the
absolute abundance of eight immune cell populations and
two stromal cell populations in tumor tissues from RNA-
seq data.
2.4. Prediction of Small Molecule Drugs. Te “limma” R
package was used to fnd diferentially expressed genes
(DEGs) between high- and low-risk groups. Ten, in order
to identify which potential target chemicals would be
helpful, we submitted the frst 1000 DEGs to the CMAP
database [19].
2.5. Statistical Analysis. R software (version 4.1.3, available
at https://www.r-project.org) was used for computational
and statistical analyses. Teir response to immunotherapy
was compared using the Wilcoxon rank sum test. Te
distinctions between the high- and low-risk categories were
ascertained using Kaplan-Meier curves and log-rank testing.
p values under 0.05 were regarded as statistically signifcant
for all analyses.
3. Results
3.1. Identifcation of Cuproptosis-Related lncRNAs in CRC
Patients. We frst analyzed the CRC mRNA dataset from the
TCGA database and obtained the expression profles of 19
CRGs and 16,876 lncRNAs. Next, we screened out 2450
CRLs by Pearson correlation analysis (|R| > 0.4, p < 0.001)
(Figure 1(a)). We further performed coexpression and
univariate Cox regression analysis and obtained 33 CRLs
that were signifcantly associated with prognosis
(Figure 1(b)). In addition, we compared the expression of
the obtained 33 CRLS in tumor tissue and normal tissue, and
the results showed that there were signifcant diferences in
their expression levels (Figures 1(c) and 1(d)).
Journal of Oncology 3

pvalue Hazard ratio

Cuproptosis
PRKAR1B–AS2 0.006 1.269 (1.071–1.503)
AL138831.2 0.003 1.943 (1.245–3.032)
AC009315.1 <0.001 1.583 (1.262–1.986)
SEPTIN7–DT 0.004 1.501 (1.141–1.976)
AC009120.5 0.005 1.418 (1.109–1.815)
NSMCE1–DT <0.001 2.508 (1.533–4.104)
AC006213.7 0.007 1.566 (1.128–2.176)
AP001619.1 0.007 1.156 (1.040–1.285)
AL122125.1 0.010 1.654 (1.131–2.421)
AC008760.1 <0.001 1.176 (1.089–1.271)
LINC01285 0.008 1.656 (1.142–2.402)
LINC01138 0.005 1.405 (1.109–1.781)
AC105460.2 0.003 1.240 (1.076–1.430)
AL034550.3 <0.001 1.827 (1.351–2.470)
PLS3–AS1 0.005 1.773 (1.188–2.648)
ZEB1–AS1 <0.001 1.226 (1.103–1.364)
ZKSCAN2–DT <0.001 1.115 (1.050–1.183)
LINC02175 0.005 1.357 (1.097–1.677)
AC022210.1 0.004 1.212 (1.064–1.382)
AC068205.2 0.008 1.274 (1.066–1.522)
LINC01410 0.002 1.266 (1.090–1.470)
AL161729.4 <0.001 1.111 (1.045–1.182)
AC007608.3 <0.001 2.391 (1.469–3.892)
AC010789.2 0.003 1.546 (1.161–2.058)
AC010463.3 0.004 1.233 (1.069–1.422)
TMEM139–AS1 0.002 12.111 (2.472–59.323)
AC083900.1 0.004 1.135 (1.042–1.237)
lncRNA AC099782.1 0.006 1.514 (1.125–2.039)
AC064836.3 0.004 1.112 (1.034–1.197)
AC010207.1 0.001 3.004 (1.539–5.861)
AC125494.3 0.008 1.734 (1.156–2.600)
Cuproptosis AC010632.1 0.002 1.937 (1.264–2.970)
AL138756.1 0.009 1.068 (1.016–1.123)
ATP7A DLST LIPT2
ATP7B FDX1 MTF1
0 10 20 30 40 50
CDKN2A GCSH NFE2L2
Hazard ratio
DBT GLS NLRP3
DLAT LIAS PDHA1
DLD LIPT1 PDHB

(a) (b)
* * *** * *** *** ** *** ** *** *** *** ** *** *** *** *** ** *** *** *** *** *** *** *** *** *** * *** * * * *** Type
AC125494.3*
AC105460.2**
AL122125.1**
30 ZEB1−AS1***
AC083900.1***
AL138831.2*
AC010632.1*
Gene expression

PRKAR1B−AS2*
AP001619.1***
20 AL138756.1*** 10
AL161729.4***
ZKSCAN2−DT***
AC068205.2*** 5
LINC01138***
10 AC064836.3***
AC008760.1***
AC022210.1*** 0
AC010463.3***
LINC02175**
AC009120.5*** −5
0 NSMCE1−DT***
AC099782.1*
TMEM139−AS1***
PRKAR1B–AS2
AL138831.2
AC009315.1
SEPTIN7–DT
AC009120.5
NSMCE1–DT
AC006213.7
AP001619.1
AL122125.1
AC008760.1
LINC01285
LINC01138
AC105460.2
AL034550.3
PLS3–AS1
ZEB1–AS1
ZKSCAN2–DT
LINC02175
AC022210.1
AC068205.2
LINC01410
AL161729.4
AC007608.3
AC010789.2
AC010463.3
TMEM139–AS1
AC083900.1
AC099782.1
AC064836.3
AC010207.1
AC125494.3
AC010632.1
AL138756.1

−10
LINC01285***
PLS3−AS1***
AC007608.3***
AC010207.1*
SEPTIN7−DT*
AL034550.3***
LINC01410***
AC010789.2***
Type AC009315.1***
AC006213.7**
Normal

Tumor
Type

Normal
Tumor

(c) (d)

Figure 1: Selection of cuproptosis-related lncRNA in CRC. (a) Te Sankey diagram shows the associations between cuproptosis-related
lncRNAs and mRNAs. (b) Te Forest plot shows 33 lncRNAs with hazard ratios (95% confdence intervals) and p-values for their as-
sociation with CRC prognosis based on univariate Cox proportional-hazards analysis. (c) Histogram of expression levels of 33 lncRNAs in
CRC tissues and paired normal tissues. (d) Heatmap of expression levels of 33 lncRNAs in CRC tissues and paired normal tissues.

3.2. Cuproptosis-Related Genotyping and GSVA Analysis. lower than that in C1 and C2, the abundance of bone
Cuproptosis is closely associated with prognosis in solid marrow dendritic cells and cytotoxic lymphocytes in C2 was
malignancies [5, 20, 21]. Based on the above hypothesis, we higher than that in C1 and C3, and there was no diference in
stratifed samples with qualitatively diferent CRC based on neutrophils and endothelial cells in C1, C2, and C3 group
the expression of 19CRGs via consensus clustering analysis. (Figures 2(c)–2(k)). Next, we compared the enrichment
Te results showed that we identifed three diferent clusters diferences of KEGG and HALLMARK signaling pathways
of modifed patterns, including 100 cases in cluster 1 (with in C1 and C3 groups by GSVA analysis, and the heat map
high CRGs and namely C1), 197 cases in cluster 2 (with showed that all pathways with statistical signifcance were
medium CRGs and namely C2), and 246 cases in cluster 3 enriched in the C1 group (Figures 3(a) and 3(b)). Tese
(with low CRGs and namely C3) (Figure 2(a) and Supple- results strongly suggest that CRGs may participate in the
mentary Figure 1). Te further survival analysis showed that immune function of CCR via multiple signaling pathways.
C1 had a worse survival advantage than C2 and C3 (Fig-
ure 2(b)). In addition, MCP counter algorithm was used to
calculate the infltration of 9 immune cells in the three 3.3. Constructing and Evaluating a Risk Score Model Based on
molecular subtypes of CRC, and the diferences were ana- CRLs in CRC. A variety of studies showed that prognostic
lyzed [22]. Te results showed that the abundance of B cells, models based on lncRNAs have guiding signifcance for
NK cells, and fbroblasts in C2 was higher than that in C3, patient prognosis [23]. Terefore, the purpose of this study is
the abundance of monocytes and CD8+T cells in C3 was to establish a CRLs-based model to facilitate the prognostic
4 Journal of Oncology

Consensus matrix 1.00 +++ 12000

+++++++++++ 0.023
++ ++++++++++
+++++ 0.13
+
++ +++++++++++++
+++++++++
++++++ ++++++++++++++++
++++
+ 0.87

Survival probability
0.75 ++++++++++++++++ +++
++++++++
+ 8000
+ ++++ +

B lineage
+ +++
++++ +
1 + ++ ++ + ++ +++ ++
0.50
0.8 4000
++ + +
0.6 0.25 +
p=0.039
0.4
0
0.2 0.00
0 0 3 6 9 12 C1 C2 C3
Time (years)
Cluster
Number at risk C1
Cluster ClusterC1 100 18 2 2 1 C2
ClusterC2 197 45 10 1 0
ClusterC3 246 67 17 9 0 C3
basis consensus silhouette
0 3 6 9 12
1 1 0.99
−0.3 Time (years)
2 2
3 3 Cluster
+ ClusterC1
+ ClusterC2
+ ClusterC3

(a) (b) (c)

8.9e−07 0.14
0.089 20 0.76
7.5 0.088 0.39
15
Neutrophils

5.0
NK cells

10

2.5
5

0.0 0

C1 C2 C3 C1 C2 C3

Cluster Cluster
C1 C1
C2 C2
C3 C3

(d) (e)
9.8e−08 4.1e−05
0.00029 5.1e−05
0.32 0.21
10
40
Monocytic lineage

CD8 T cells

20 5

0 0

C1 C2 C3 C1 C2 C3

Cluster Cluster
C1 C1
C2 C2
C3 C3

(f ) (g)
Figure 2: Continued.
Journal of Oncology 5

1.6e−12 8.3e−12
15 0.0054 0.024
15

Myeloid dendritic cells

Cytotoxic lymphocytes
0.0055 0.00042
10 10

5 5

0 0

C1 C2 C3 C1 C2 C3
Cluster
Cluster C1
C1 C2
C2 C3
C3
(h) (i)
0.45 0.00083
30 0.11 0.014
0.39 0.78
2000
Endothelial cells

20

Fibroblasts
1000
10

0 0

C1 C2 C3 C1 C2 C3
Cluster Cluster
C1 C1
C2 C2
C3 C3

(j) (k)

Figure 2: Consensus clustering and the diferent immune profles between tree clusters. (a) Consensus map of NMF clustering. (b) Overall
survival curves of the three molecular subtypes. (c–k) Comparisons of the abundance of infltrating immune function between C1, C2 and
C3.

prediction of CRC. We previously obtained 33 CRLs with divided the integrated cohort into low-risk and high-risk
prognostic values, which were further screened by LASSO groups based on the median risk score (Figures 4(e) and
regression. A total of 6 lncRNAs were obtained, and their 4(f )). Te above results indicated that the risk score model
risk coefcients were calculated. Specifcally, the risk score based on CRLs has a good predictive efciency for the
model for predicting CRC prognosis based on 6 CRLs is prognosis of CRC patients.
shown as follows: risk score � expression value of
AC009315.1 ∗ 0.15835365544805 + expression value of
3.4. Te Correlation Analysis between the Clinical Features and
PLS3-AS1 ∗ 0.100587632623172 + expression value of ZEB1-
CRLs Risk Score Model for CRC Patients. We previously
AS1 ∗ 0.0274302502732273 + expression value of
established a risk score model based on CRLs and found that
AC007608.3 ∗ 0.0549982300165668 + expression value of
it could accurately predict the survival prognosis of CRC
AC010789.2 ∗ 0.166645095217608 + expression value of
patients. To further explore the value of this model, we
AC010207.1 ∗ 0.403357464363707. Next, we randomly di-
analyzed the correlation of this model with the clinical
vided the CRC cohort patients into two cohorts, a training
features (age, gender and stage) of CRC patients
cohort (n � 382) and a validation cohort (n � 161), in a 7 : 3
(Figure 5(a)). Te results showed that the low-risk group had
ratio. In the training and validation cohort, the patients were
better survival prognosis in patients aged >65 years, male,
divided into low-risk and high-risk groups based on the
stage I-II, stage III-IV, T3-4, N1-2, and M0. Tere was no
median risk score. Surprisingly, we found that the low-risk
diference in survival prognosis between high and low-risk
group had a higher survival rate than the high-risk group,
groups in patients aged ≤65 years, female, T1-2, N0, and M1
both in the training and validation cohort (Figures 4(a) and
(Figures 5(b)–5(d)).
4(b)). Te AUC values of the 1-year, 3-year and 5-year ROC
curves of the training cohort were 0.756, 0.737, and 0.649,
respectively, while the AUC values of the 1-year, 3-year and 3.5. Enrichment and Drug Sensitivity Analysis of CRLs Risk
5-year ROC curves of the validation cohort were 0.688, Score Model. In order to clarify the specifc mechanism of
0.652, and 0.728, respectively, (Figures 4(c) and 4(d)). In the CRLs risk score in CCR, we further analyzed the po-
addition, in the training and validation cohort, the signature tential functional pathway of the high-risk and low-risk
6 Journal of Oncology

Cluster 2
KEGG_ABC_TRANSPORTERS 1

KEGG_CIRCADIAN_RHYTHM_MAMMAL 0
-1
KEGG_TASTE_TRANSDUCTION
-2
KEGG_LINOLEIC_ACID_METABOLISM

KEGG_GLYCOSPHINGOLIPID_BIOSYNTHESIS_GLOBO_SERIES

KEGG_SYSTEMIC_LUPUS_ERYTHEMATOSUS

KEGG_ANTIGEN_PROCESSING_AND_PRESENTATION

KEGG_ASTHMA

KEGG_GLUTATHIONE_METABOLISM

KEGG_FOLATE_BIOSYNTHESIS

KEGG_PROTEIN_EXPORT

KEGG_NUCLEOTIDE_EXCISION_REPAIR

KEGG_HOMOLOGOUS_RECOMBINATION

KEGG_DNA_REPLICATION

KEGG_MISMATCH_REPAIR

KEGG_PROTEASOME

KEGG_PYRIMIDINE_METABOLISM

KEGG_RNA_POLYMERASE

Cluster
C1

C3

(a)
Cluster
1

0
HALLMARK_OXIDATIVE_PHOSPHORYLATION -1

-2

HALLMARK_MYC_TARGETS_V1

HALLMARK_E2F_TARGETS

HALLMARK_MTORC1_SIGNALING

HALLMARK_DNA_REPAIR

Cluster
C1

C3

(b)

Figure 3: Genomic variation analysis of pathways in C1 and C3 cluster. (a) Heatmap showing the diferentially expressed pathways between
C1 and C3 cluster based on KEGG pathway enrichment analysis. (b) Heatmap showing the diferentially expressed pathways between C1
and C3 cluster based on HALLMARK pathway enrichment analysis.
Journal of Oncology 7

1.00 1.00 +++++++++

++ ++++++++++++
+++ +++++++++
++++++++ ++++++++
Survival probability

Survival probability
0.75 0.75 ++++++++ + ++ +
++
++ + + + +
++
0.50 0.50

0.25 0.25 + +
p=0.003 p<0.001

0.00 0.00
0 3 6 9 12 0 3 6 9 12
Number at risk Time (years) Time (years)
Number at risk
High risk 191 41 9 4 1 76 15 2 0 0
Risk

High risk

Risk
Low risk 191 44 9 4 0 Low risk 85 30 9 4 0
0 3 6 9 12 0 3 6 9 12
Time (years) Time (years)
Risk
Risk
High risk
+ High risk
Low risk + Low risk

(a) (b)
1.0 1.0

0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2

0.0 0.0

0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
1 – specificity 1 – specificity

AUC at 1 years: 0.756 AUC at 1 years: 0.688

AUC at 3 years: 0.737 AUC at 3 years: 0.652
AUC at 5 years: 0.649 AUC at 5 years: 0.728

(c) (d)
6
High risk 5 High risk
5 Low Risk Low Risk
Risk score

Risk score

4
4
3 3
2 2
1 1
0 100 200 300 0 50 100 150
Patients (increasing risk socre) Patients (increasing risk socre)
12 Dead 10 Dead
Survival time

10 Alive
Survival time

Alive 8
(years)

8
(years)

6
6
4 4
2 2
0 0
0 100 200 300 0 50 100 150
Patients (increasing risk socre) Patients (increasing risk socre)

(e) (f )

Figure 4: Establishment and validation of a six-gene risk score model (a), (b) Kaplan-Meier curves of overall survival for the high and low-
risk groups in the training and validation datasets. (c), (d) Time-dependent receiver operating characteristic curves for the risk score in the
training and validation datasets. (e), (f ) Distribution of risk scores and survival status for the training and validation datasets.
8 Journal of Oncology

Risk
Age
Gender
Stage*
T
N*
M**
Risk Age Gender Stage* T N* M**
high <=65 FEMALE I T1 N0 M0
low >65 MALE II T2 N1 M1
III T3 N2
IV T4
(a)
Patients with <=65
1.00
Survival probability

0.75

0.50

0.25
p = 0.33
0.00
0 3 6 9 12
Time (years)
Number at risk
high 114 26 5 2 1
Risk

low 116 29 7 4 0
0 3 6 9 12
Time (years)

Risk
high
low
(A)
Patients with >65
1.00
Survival probability

0.75

0.50

0.25
p < 0.0001
0.00
0 3 6 9 12
Time (years)
Number at risk
high 146 29 6 2 0
Risk

low 154 45 11 4 0
0 3 6 9 12
Time (years)
Risk
high
low
(B)

(b)
Figure 5: Continued.
Journal of Oncology 9

Patients with MALE Patients with I−II

1.00 1.00

Survival probability

Survival probability
0.75 0.75

0.50 0.50

0.25 0.25
p < 0.0001 p = 0.036
0.00 0.00
0 3 6 9 12 0 3 6 9 12
Time (years) Time (years)
Number at risk Number at risk
high 141 28 4 1 0 high 133 34 5 2 0
Risk

Risk
low 139 43 9 2 0 low 169 52 13 4 0
0 3 6 9 12 0 3 6 9 12
Time (years) Time (years)
Risk Risk
high high
low low
(A) (A)
Patients with FEMALE
Patients with III−IV
1.00
1.00
Survival probability

0.75

Survival probability
0.75

0.50
0.50

0.25
p = 0.087 0.25
p = 0.0033
0.00
0.00
0 3 6 9 12
0 3 6 9 12
Time (years)
Time (years)
Number at risk
Number at risk
high 119 27 7 3 1
Risk

high 127 21 6 2 1
Risk
low 131 31 9 6 0
low 101 22 5 4 0
0 3 6 9 12
0 3 6 9 12
Time (years)
Time (years)
Risk
Risk
high
high
low
low
(B)
(B)

(c) (d)

Figure 5: Te correlation analysis between the clinical features and CRLs risk score model for CRC patients. (a) Clinicopathological
information of CRC patients is arranged by increasing risk score. (b) KM survival curve of high and low risk groups in patients with age ≤65
(A) and age >65 (B). (c) KM survival curves of high and low risk groups in male (A) and female (B) patients. (d) KM survival curves of high
and low risk groups for stage I-II (A) and III-IV (B) patients.

groups. Te results showed that the diferentially expressed
genes in the high-risk group and the low-risk group were
mainly enriched in multiple signaling pathways, such as
DNA packaging, chromatin assembly and neutrophil ex-
tracellular trap formation (Figures 6(a) and 6(b)). In ad-
dition, we further analyzed the association between the CRLs
risk score and the efcacy of chemotherapy in the treatment
of CRC. It showed that the high-risk group was associated
with lower half inhibitory centration (IC50) of chemo-
therapeutic drugs, such as AZ8055, Paclitaxel, and AKT
inhibitor VII, while the IC50 of Cisplatin, 5-Fu, and Tra-
metinib was higher (Figures 6(c)–6(h)). Te results showed
that the CRLs risk score model could be used as a predictor
of chemical sensitivity in the future.
3.6. Te Relationship between TME and CRLs Risk Score in
CRC. Te immune microenvironment of tumors is closely
related to tumor progression. Tumor cells interact with
immune cells, thereby inhibiting the function of immune
cells and fnally leading to tumor immune escape [24, 25].
Terefore, we continued to investigate whether the CCR
immune microenvironment was associated with CRLs risk
scores. We assessed the immune microenvironment of CRC
by the ESTIMATE algorithm and observed the diferences in
the stromal score and immune score between the high-risk
group and the low-risk group. As shown in Figure 7(a),
lower immune scores were exhibited in the high-risk group.
Te distribution of 22 immune cells in CRC patients is
shown in Figure 7(b). Next, we further calculated the in-
fltration abundance of immune cells by the CIBORESORT
algorithm. Te results showed that the infltrating abun-
dance of CD8+ T cells in the low-risk group was higher than
that in the high-risk group (Figure 7(c)). Moreover, the
boxplot of immune function analysis showed that the scores
of chemokine receptors, HLA and MHC in the high-risk
group were signifcantly lower than those in the low-risk
group (Figure 7(d)). Immune checkpoints are important
predictors for assessing immunotherapy response [26].
Terefore, we evaluated the association of 12 immune
checkpoints with CRLs. As shown in Figure 7(e), all
10 Journal of Oncology

protein heterodimerization activity

protein−DNA complex Systemic lupus
DNA packaging complex erythematosus
nucleosome
DNA conformation change
Alcoholism qvalue
chromatin remodeling
protein−DNA complex 0.001
Term

subunit organization 0.002

protein−DNA complex assembly Neutrophil 0.003
0.004
extracellular
DNA packaging
trap formation
chromatin assembly or disassembly
nucleosome organization
Viral carcinogenesis
chromatin assembly
nucleosome assembly
0 2 4 6 0 2 4 6
Count Count

ONTOLOGY
BP
CC
MF
(a) (b)
0.0092 0.0072

AZD8055 senstivity (IC50)

Cisplatin senstivity (IC50)

2
4
0
2
−2
0
low high low high
Risk Risk

Risk Risk
low low
high high
(c) (d)
0.00036 0.015
7.5
Paclitaxel senstivity (IC50)

5−Fluorouracil senstivity

0
5.0
−2
(IC50)

2.5
−4
0.0

−6 −2.5
low high low high
Risk Risk

Risk Risk
low low
high high

(e) (f )
Figure 6: Continued.
Journal of Oncology 11

0.0011 0.0011

Trametinib senstivity (IC50)

4

AKT inhibitor VIII

4

senstivity (IC50)
3

2 0

1 −4

0 −8
low high low high
Risk Risk

Risk Risk
low low
high high
(g) (h)

Figure 6: GO and KEGG enrichment and drug sensitivity analysis between high-risk and low-risk groups. (a) GO analysis for the high-risk
group and low-risk group. (b) KEGG analysis for the high-risk group and low-risk group. (c) Diferences in chemosensitivity to cisplatin in
high and low risk groups. (d) Diferences in chemosensitivity to AZD8055 in high and low risk groups. (e) Diferences in chemosensitivity to
paclitaxel in high and low risk groups. (f ) Diferences in chemosensitivity to 5-Fluorouracil in high and low risk groups. (g) Diferences in
sensitivity to AKT inhibitor III treatment between high and low-risk groups. (h) Diferences in sensitivity to trametinib treatment between
high- and low-risk groups. Sensitivity to chemotherapeutic drugs is expressed by the half inhibitory centration (IC50) of
chemotherapeutic drugs.

lncRNAs in the CRLs risk model were positively associated
with most immune checkpoints. Finally, we analyzed the
relationship of four common immune checkpoints with risk
scores, and the results showed that CD274 (PD-L1), CTLA4,
and HAVCR2 (TIM3) were positively associated with risk
scores (Figure 7(f )). Te above data strongly suggested that
CRGs play an important role in the regulation of the CCR
immune microenvironment.
3.7. Correlataion between Immunotherapy Responsiveness
and CRLs Risk Score. MSI is an important indicator for
evaluating the efcacy of immunotherapy in CRC [27].
Terefore, we explored the association of MSS, MSI-L and
MSI-H with CRLS scores. Te result showed that MSI-H
patients had lower risk scores than MSI-L patients
(Figure 8(a)). In recent years, IPS and TIDE have been
widely used to evaluate the efcacy of immunotherapy
[28, 29]. Our analysis revealed that IPS scores were signif-
icantly higher in the low CRLs score group (Figures 8(b) and
8(c)). Consistently, the low-risk group had a lower immune
exclusion, immune dysfunction and TIDE scores than the
high-risk group (Figures 8(d)–8(f )). Tese fndings in-
directly suggest that CRLs may play a key role in mediating
immune responses in CRC.
4. Discussion
In recent years, the gradual increase in the incidence of CRC
has attracted many researchers to lucubrate its occurrence,
development and treatment. Te resistance of tumors to
antitumor therapy has made people gradually realize the
importance of programmed cell death, such as autophagy,
pyroptosis and ferroptosis [30–32]. Cuproptosis is a newly
discovered type of cell death that can be induced by a variety
of drugs [33]. Terefore, a full understanding of the specifc
mechanisms of cuproptosis is critical to guide the treatment
of CCR.
In this study, frstly, we found that CRGs were closely
associated with CCR immune cell infltration. Next, we
identifed 6 CRLs signifcantly associated with prognostic
and then developed a prognostic model. In addition, we
validate the accuracy of the CRLs model and initially explore
its underlying mechanisms. Finally, we evaluated the re-
lationship between risk scores and immune cell infltration,
drug sensitivity, and immunotherapy efcacy.
In the past decade, more and more studies attempted to
establish lncRNA-based prognostic models in order to
provide guidance for the prognosis of various malignant
tumors. Tang et al. analyzed the expression of ferroptosis-
related lncRNAs in head and neck squamous cell carcinoma
in a public database, constructed a prognostic model, and
further confrmed that it has a good predictive efect. Te
AUC area for 1 year, 3 years, and 5 years is 0.78, 0.83, and
0.71, respectively [34]. Song et al. analyzed the expression of
pyroptosis-related lncRNAs in lung cancer tissues and
constructed a prognostic model with good predictive ability.
Te AUC area for 1 year, 3 years and 5 years is 0.757, 0.728,
and 0.685, respectively [35]. In this study, the areas under the
AUC curve of our prognostic model at 1 year, 3 years, and
5 years were 0.756, 0.737, and 0.649, respectively. Compared
with previous studies, this model shows no weak predictive
ability and has good clinical application value.
Te immune microenvironment of tumors is regulated
by a variety of cells, including tumor cells themselves, im-
mune cells, and fbroblasts [36]. Among them, immune cells
play a major role in regulating the tumor immune micro-
environment [37]. In recent years, eforts have been made to
explore new approaches to treat CCR. Te advent of im-
munotherapy has brought new hope to this idea. A variety of
 12

Fraction
TME score

0.0
0.2
0.4
0.6
B cells naive

−4000
−2000
0
2000
4000
6000

Risk
B cells memory

**

low
high
Plasma cells

Risk
T cells CD8

**
low
T cells CD4 naive

high
T cells CD4 memory resting StromalScore
T cells CD4 memory activated
T cells follicular helper
T cells regulatory (Tregs)

(c)
(a)
T cells gamma delta
NK cells resting
*

NK cells activated ImmuneScore

Monocytes

*
Macrophages M0
Macrophages M1
Macrophages M2
Dendritic cells resting
Dendritic cells activated ESTIMATEScore
Mast cells resting
Mast cells activated
Eosinophils
Neutrophils

Relative Percent (%)

0
20
40
60
80
100

Score

Figure 7: Continued.
0.00
0.25
0.50
0.75
1.00

APC_co_inhibition

Risk
low
high
APC_co_stimulation
Low risk

CCR
Check−point
Cytolytic_activity
*
(b)

(d)
HLA
Inflammation−promoting
*

MHC_class_I
High risk

Parainflammation
T_cell_co−inhibition
T_cell_co−stimulation
Type_I_IFN_Reponse
Monocytes
T cells CD8

Eosinophils
Neutrophils
Plasma cells
B cells naive
B cells memory

NK cells resting

Type_II_IFN_Reponse
Mast cells resting
Macrophages M2
Macrophages M1
Macrophages M0
T cells CD4 naive

NK cells activated

Mast cells activated

T cells gamma delta

Dendritic cells resting

T cells follicular helper

Dendritic cells activated

T cells regulatory (Tregs)
T cells CD4 memory resting
T cells CD4 memory activated
Journal of Oncology
Journal of Oncology 13

VSIR ***
6
TIGIT *** * ** R = 0.11, p = 0.0091
R = 0.072, p = 0.096 6

PDCD1 expression

CD274 expression
PDCD1 * *** p<0.001 4
4
LAG3 * ** p<0.01
KLRD1 ** *** * *** * p<0.05 2 2
IDO1 ** ** * Correlation
0.2 0 0
IAPP *** ** *** * *** 0.1 1 2 3 4 5 6 1 2 3 4
HAVCR2 ** *** *** 0.0 Risk score
Risk score
CTLA4 *** ** * **
CD96 ** *** ***
R = 0.15, p = 0.00039
R = 0.13, p = 0.002

HAVCR2 expression
5

CTLA4 expression
CD274 *** *** ** 4
CD226 * *** ** *** 4
3
3
AC009315.1

PLS3−AS1

ZEB1−AS1

AC007608.3

AC010789.2

AC010207.1

riskScore
2
2
1
1
0
1 2 3 4 5 6 1 2 3 4 5 6
Risk score Risk score

(e) (f )

Figure 7: Te relationship between the immune landscape and CRLS in CRC. (a) Diferences in ESTIMATE score, immune score, and
stromal score between high- and low-risk groups. (b), (c) Landscape of immune cell infltration in high-risk and low-risk groups. (d)
Diferences in immune cell scores between high-risk and low-risk groups in the CRLS risk score model. (e) Correlation analysis of immune
checkpoints and 6 lncRNA expression profles. (f ) Relationship of risk score with PDCD1, CD274, CTLA4, and HAVCR2 expression.

0.0071
3.0 0.1
0.034 0.00041 3.5e−05
ips_ctla4_neg_pd1_pos

ips_ctla4_pos_pd1_neg

10 10
2.5
Risk score

9
2.0 8
8

1.5 6 7
6
1.0
MSS MSI−L MSI−H low high low high
Riskscore Riskscore
MSI
MSS Risk Risk
MSI−L low low
MSI−H high high
(a) (b) (c)

* 3 *
ns
2 2
2
Dysfunction

1
1
TIDE
Exclusion

0 0 0
−1
−1
−2 −2
−3 −2

Low−risk High−risk Low−risk High−risk Low−risk High−risk

Risk Risk Risk

Low−risk Low−risk Low−risk
High−risk High−risk High−risk

(d) (e) (f)

Figure 8: Correlation between immunotherapy responsiveness and CRLs risk score model. (a) Association of MSI status and risk score. (b)
IPS scores between high and low CRLs score groups when CTLA-4 negative and PD1 positive. (c) IPS scores between high and low CRLs
score groups when CTLA-4 positive and PD1 negative. (d-f ) Te diferences in immune exclusion, immune dysfunction, and TIDE scores
between high-risk and low-risk groups.
14
evidence indicates that infltrating lymphocytes play an
important role in the prognosis of various solid tumors and
have potential predictive value [38]. In this study, we found
that diferent groups of CRGs have diferences in the dis-
tribution of immune cells, which indirectly suggests the
existence of a relationship between CRGs and the immune
microenvironment. More interestingly, we constructed
a prognostic model based on CRLs and also showed an
association between the risk score and the proportion of
immune cells in the tumor microenvironment. Tis result
further supported the relationship between cuproptosis and
the immune microenvironment of CCR. However, its spe-
cifc mechanism needs to be further studied in the future.
Cuproptosis is a novel mode of cell death for which
research is currently rather limited. In this study, we found
that there were signifcant diferences in the infltration of
various immune cells under diferent patterns of CRLs. More
interestingly, the risk scores of the prognostic models
constructed based on CRLs were also signifcantly diferent
from the immune microenvironment of CRC and its
multiple immune checkpoints. Tese data strongly sug-
gested that there is a strong interrelationship prior to
cuproptosis and immunity. Given the existence of an ex-
tremely complex network of molecular interactions within
cells. In addition, there are some unsatisfactory aspects of
this study. Firstly, all data in this study were obtained from
public databases, lacking further support from clinical data.
Secondly, the mechanism by which the CRLs model regu-
lates the immune microenvironment has not been thor-
oughly investigated. Tese issues deserve further research in
the future.
In conclusion, in this study, we revealed multiple roles of
CRGs and CRLs in CCR. Firstly, CRGs were closely related
to CCR immune cell infltration. Secondly, the risk scoring
model based on CRLs has a good predictive ability for the
overall survival of CCR. In addition, the risk score of CRLs
might have potential guiding value for the application of
various antitumor drugs. Moreover, the risk score of CRLs
was closely related to the immune cell infltration of CCR.
Finally, the CRLs risk model might have potential instructive
value for immunotherapy.
Abbreviations
CCR: Colorectal cancer
CRGs: Cuproptosis-related genes
CRLs: Cuproptosis-related lncRNA
KEGG: Kyoto encyclopedia of genes and genomes
TCGA: Te cancer genome atlas.
Data Availability
Te data and result in this study are available from the
corresponding author for reasonable request.
Ethical Approval
Te study was conducted in accordance with the Declaration
of Helsinki, and the protocol was approved by the Ethics
Journal of Oncology
Committee of Te First Afliated Hospital of Jinan
University.
Conflicts of Interest
Te authors declare that they have no conficts of interest.
Authors’ Contributions
Yi Yang, Xiaoli Wang, and Jin Lu contributed equally to this
work. All authors read and approved the fnal manuscript.
Acknowledgments
Te authors would like to express our sincere thanks for
sharing the data from the Cancer Genome Atlas (TCGA)
database. Tis research was supported by the following
funding agency: National Natural Science Foundation of
China (grant no. 82002491); the Natural Science
Foundation of Guangdong Province of China (grant no.
2020A1515110049).
Supplementary Materials
Supplementary Figure 1. Unsupervised clustering of 19
cuproptosis genes in the TCGA cohort. Supplementary
Figure 2. Identifcation of the prognostic genes based on
CRLs by LASSO regression. Supplementary Figure 3. Te
correlation between the clinical features and risk score of
CRC patients. (Supplementary Materials)
References
[1] L. H. Biller, D. H. Schrag, and D. Schrag, “Diagnosis and
Diagnosis and Treatment of Metastatic Colorectal Cancer-
reatment of metastatic colorectal cancer: a review,” Jama,
vol. 325, no. 7, pp. 669–685, 2021.
[2] B. Dariya, S. Aliya, G. P. Nagaraju, A. Alam, and N. Merchant,
“Colorectal Colorectal Cancer Biology, Diagnosis, and
Terapeutic Approachesancer biology, diagnosis, and thera-
peutic approaches,” Critical Reviews in Oncogenesis, vol. 25,
no. 2, pp. 71–94, 2020.
[3] T. Yamamoto, K. Kawada, and K. Obama, “Infammation-
Infammation-Related Biomarkers for the Prediction of
Prognosis in Colorectal Cancer Patientselated biomarkers for
the prediction of prognosis in colorectal cancer patients,”
International Journal of Molecular Sciences, vol. 22, no. 15,
p. 8002, 2021.
[4] D. Bruni, H. K. Angell, and J. K. Galon, “Te immune
contexture and Immunoscore in cancer prognosis and
therapeutic efcacy,” Nature Reviews Cancer, vol. 20, no. 11,
pp. 662–680, 2020.
[5] P. Tsvetkov, R. D. Spangler, R. Humeidi et al., “Copper in-
duces cell death by targeting lipoylated TCA cycle proteins,”
Science, vol. 375, no. 6586, pp. 1254–1261, 2022.
[6] M. V. Babak and D. V. Ahn, “Modulation of Modulation of
Intracellular Copper Levels as the Mechanism of Action of
Anticancer Copper Complexes: Clinical Relevancentracellular
copper levels as the mechanism of action of anticancer copper
complexes: clinical relevance,” Biomedicines, vol. 9, no. 8,
p. 852, 2021.
Journal of Oncology
[7] S. Ishida, P. Andreux, D Hanahan, J. Auwerx, and C. Poitry-
Yamate, “Bioavailable copper modulates oxidative phos-
phorylation and growth of tumors,” Proceedings of the Na-
tional Academy of Sciences, vol. 110, no. 48, pp. 19507–19512,
2013.
[8] E. J. Ge, S. Lutsenko, V. M. Gohil et al., “Connecting copper
and cancer: from transition metal signalling to metal-
loplasia,” Nature Reviews Cancer, vol. 22, no. 2, pp. 102–113,
2022.
[9] Z. Bian, R. Fan, and L. Xie, “A A Novel Cuproptosis-Related
Prognostic Gene Signature and Validation of Diferential
Expression in Clear Cell Renal Cell Carcinomaovel
cuproptosis-related prognostic gene signature and validation
of diferential expression in clear cell renal cell carcinoma,”
Genes, vol. 13, no. 5, p. 851, 2022.
[10] N. A. Johdi and N. F. Sukor, “Colorectal Colorectal Cancer
Immunotherapy: Options and Strategiesancer immunother-
apy: options and strategies,” Frontiers in Immunology, vol. 11,
p. 1624, 2020.
[11] Z. Payandeh, S. Khalili, M. H. Somi et al., “PD-1/PD-L1-
dependent immune response in colorectal cancer,” Journal of
Cellular Physiology, vol. 235, no. 7-8, pp. 5461–5475, 2020.
[12] S. Djebali, F. Schlesinger, A. Tanzer et al., “Landscape of
transcription in human cells,” Nature, vol. 489, no. 7414,
pp. 101–108, 2012.
[13] A. Sanchez Calle, Y. Kawamura, T. Ochiya, F. Takeshita, and
Y. Yamamoto, “Emerging roles of long non-coding RNA in
cancer,” Cancer Science, vol. 109, no. 7, pp. 2093–2100, 2018.
[14] A. Bhan, M. Soleimani, and S S. Mandal, “Long Long Non-
coding RNA and Cancer: A New Paradigmoncoding RNA
and cancer: a new paradigm,” Cancer Research, vol. 77, no. 15,
pp. 3965–3981, 2017.
[15] J. You, J. Li, Q. Li et al., “Oncogenic long intervening non-
coding RNA Linc00284 promotes c-Met expression by
sponging miR-27a in colorectal cancer,” Oncogene, vol. 40,
no. 24, pp. 4151–4166, 2021.
[16] X. Li, F. Jin, and Y. Li, “A novel autophagy-related lncRNA
prognostic risk model for breast cancer,” Journal of Cellular
and Molecular Medicine, vol. 25, no. 1, pp. 4–14, 2021.
[17] H. Yuan, T. Yang, S. Fan et al., “Prognostic Prognostic Risk
Model and Tumor Immune Environment Modulation of
m5C-Related LncRNAs in Pancreatic Ductal Adenocarcino-
maisk model and tumor immune environment modulation of
m5C-related LncRNAs in pancreatic ductal adenocarci-
noma,” Frontiers in Immunology, vol. 12, Article ID 800268,
2021.
[18] B. Chen, M. S. Khodadoust, C. L. Liu et al., “Profling Profling
Tumor Infltrating Immune Cells with CIBERSORTumor
infltrating immune cells with CIBERSORT,” Methods in
Molecular Biology, vol. 1711, pp. 243–259, 2018.
[19] H. Li, W. Chen, H. Pan et al., “CMap analysis identifes
Atractyloside as a potential drug candidate for type 2 diabetes
based on integration of metabolomics and transcriptomics,”
Journal of Cellular and Molecular Medicine, vol. 24, no. 13,
pp. 7417–7426, 2020.
[20] D. Tang, X. Chen, G. Kroemer, and G. Kroemer, “Cuprop-
tosis: a copper-triggered modality of mitochondrial cell
death,” Cell Research, vol. 32, no. 5, pp. 417-418, 2022.
[21] Y. Jiang, Z. Huo, Z. Wu et al., “Copper-induced tumor cell
death mechanisms and antitumor theragnostic applications of
copper complexes,” Nanomedicine, vol. 17, no. 5, pp. 303–324,
2022.
15
[22] P. Ahluwalia, M. Ahluwalia, A. K. Mondal et al., “Immu-
nogenomic Immunogenomic Gene Signature of Cell-Death
Associated Genes with Prognostic Implications in Lung
Cancerene signature of cell-death associated genes with
prognostic implications in lung cancer,” Cancers, vol. 13,
no. 1, p. 155, 2021.
[23] S. Xu, Y. Zhou, J. Xie et al., “Integrated Integrated Analysis of
a Ferroptosis-Related LncRNA Signature for Evaluating the
Prognosis of Patients with Colorectal Cancernalysis of a fer-
roptosis-related LncRNA signature for evaluating the prog-
nosis of patients with colorectal cancer,” Genes, vol. 13, no. 6,
p. 1094, 2022.
[24] X. Jiang, X. Li, M. Zhou et al., “Role of the tumor micro-
environment in PD-L1/PD-1-mediated tumor immune es-
cape,” Molecular Cancer, vol. 18, no. 1, p. 10, 2019.
[25] A. Simiczyjew, E. Dratkiewicz, R. Matkowski et al., “Te Te
Infuence of Tumor Microenvironment on Immune Escape of
Melanomanfuence of tumor microenvironment on immune
escape of melanoma,” International Journal of Molecular
Sciences, vol. 21, no. 21, p. 8359, 2020.
[26] G. Morad, B. A. Helmink, P. Sharma, J. A. A. Wargo,
P. Sharma, and J. A. Wargo, “Hallmarks of response, re-
sistance, and toxicity to immune checkpoint blockade,” Cell,
vol. 184, no. 21, pp. 5309–5337, 2021.
[27] K. Ganesh, Z. K. Stadler, L A. Diaz et al., “Immunotherapy in
colorectal cancer: rationale, challenges and potential,” Nature
Reviews Gastroenterology & Hepatology, vol. 16, no. 6,
pp. 361–375, 2019.
[28] J. Wu, L. Li, H. Zhang et al., “A risk model developed based on
tumor microenvironment predicts overall survival and as-
sociates with tumor immunity of patients with lung adeno-
carcinoma,” Oncogene, vol. 40, no. 26, pp. 4413–4424, 2021.
[29] C. P. Gui, J. H. Wei, Y. H Chen et al., “A new thinking:
extended application of genomic selection to screen mul-
tiomics data for development of novel hypoxia-immune
biomarkers and target therapy of clear cell renal cell carci-
noma,” Briefngs in Bioinformatics, vol. 22, no. 6, Article ID
bbab173, 2021.
[30] A. V. Onorati, M. Dyczynski, R. Ojha et al., “Targeting
autophagy in cancer,” Cancer, vol. 124, no. 16, pp. 3307–3318,
2018.
[31] P. Yu, X. Zhang, C. Peng et al., “Pyroptosis: mechanisms and
diseases,” Signal Transduction and Targeted Terapy, vol. 6,
no. 1, p. 128, 2021.
[32] C. Liang, X. Zhang, M. Yang, X. Dong, M. Yang, and X. Dong,
“Recent Recent Progress in Ferroptosis Inducers for Cancer
Terapyrogress in ferroptosis inducers for cancer therapy,”
Advanced Materials, vol. 31, no. 51, Article ID 1904197, 2019.
[33] Y. Wang, L. Zhang, F. Zhou, and F. Zhou, “Cuproptosis: a new
form of programmed cell death,” Cellular & Molecular Im-
munology, vol. 19, no. 8, pp. 867-868, 2022.
[34] Y. Tang, C. Li, Y. J. Zhang, Z. H. Wu, Y. J. Zhang, and
Z. H. Wu, “Ferroptosis-Related Long Non-Coding RNA
signature predicts the prognosis of Head and neck squamous
cell carcinoma,” International Journal of Biological Sciences,
vol. 17, no. 3, pp. 702–711, 2021.
[35] J. Song, Y. Sun, L. Xi et al., “A novel pyroptosis-related
lncRNA signature for prognostic prediction in patients
with lung adenocarcinoma,” Bioengineered, vol. 12, no. 1,
pp. 5932–5949, 2021.
[36] H. Locy, S. Mey, K. Tielemans et al., “Immunomodulation of
the Immunomodulation of the Tumor Microenvironment:
16 Journal of Oncology

Turn Foe Into Friendumor microenvironment: turn foe into

friend,” Frontiers in Immunology, vol. 9, p. 2909, 2018.
[37] T. F Gajewski, H. Schreiber, Y. X. Fu, H. Schreiber, and
Y. X. Fu, “Innate and adaptive immune cells in the tumor
microenvironment,” Nature Immunology, vol. 14, no. 10,
pp. 1014–1022, 2013.
[38] F. Petitprez, M. Meylan, A. de Reyniès et al., “Te Te Tumor
Microenvironment in the Response to Immune Checkpoint
Blockade Terapiesumor microenvironment in the response
to immune checkpoint blockade therapies,” Frontiers in
Immunology, vol. 11, p. 784, 2020.