The Prognostic Value and Immunological Role of CD44 in Pan-Cancer Study

doi:10.21203/rs.3.rs-2109334/v1

Download PDF

Case Report

The Prognostic Value and Immunological Role of CD44 in Pan-Cancer Study

https://doi.org/10.21203/rs.3.rs-2109334/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

To investigate the correlation between cluster of differentiation-44 (CD44) expression and immunotherapy response and identify its possible predictive value in pan-cancer.

Methods

Datasets of 33 cancer types from The Cancer Genome Atlas (TCGA) database were applied to investigate the relationship of CD44 expression with prognosis, tumor mutational burden (TMB), and microsatellite instability (MSI), and determine its potential prognostic value in pan-cancer. Patients were split into high-risk and low-risk cancer groups based on the survival outcomes of various cancer types. Additionally, the underlying mechanisms of CD44 in the tumor microenvironment (TME) were analyzed using ESTIMATE and CIBERSORT algorithms and Gene Set Enrichment Analysis (GSEA).

Results

Variable expression levels of CD44 between tumor and adjacent normal tissues were identified in pan-cancer datasets, and the survival analysis revealed that CD44 expression was associated with multiple clinical annotations and survival indicators in several cancers. Besides, the expression of CD44 was significantly associated with TMB and MSI in some types of cancer.

Conclusion

Our study demonstrated that CD44 shows great promise as a prognostic biomarker in numerous cancers, which will assist in developing new strategies in immunotherapy.

CD44

prognosis

pan-cancer

biomarker

tumor microenvironment

Cancer morbidity and mortality are rapidly escalating, especially in developed countries¹. According to a systematic analysis of 29 cancer types in 195 countries, about 9.6 million individuals died from cancer-related deaths in 2017². Currently, cancer therapies include surgery, chemotherapy, targeted therapy, and radiotherapy. Despite the clinical success of these treatments, the prognosis and tolerability of cancer patients remain unsatisfactory owing to the limited efficacy and side effects of anti-cancer agents³. Fortunately, cancer immunotherapy has revolutionized cancer management following advancements in immunotherapy, presenting encouraging results in clinical trials⁴. However, non-responsiveness to immunotherapy was also observed in specific cancer types and patients, indicating the presence of intrinsic resistance or naturally acquired resistance⁵. Hence, it is of tremendous significance to ascertain novel biomarkers for cancer diagnosis and, most importantly, predict the efficacy of immunotherapy^6,7.

Cluster of differentiation-44 (CD44) is a member of the non-kinase, single-span transmembrane glycoproteins family, which can contribute to cancer stem cells (CSCs) function and is generally recognized as a molecular marker for CSCs ⁸. It has been well documented that CD44 expression in CSCs is associated with metastasis and capacity resisting to apoptosis of cancer cells ⁹. Similarly, Zhang et al. ¹⁰ reported that CD44 was related to cellular states and phenotypes of tumour cells in breast cancer, while Gomez et al.¹¹ noted that CD44 expression influenced cancer cell plasticity through tumour-associated macrophages (TAM), inferring that CD44 is a surface marker defining HNSC. The association of CD44 expression with other cancers, including prostate cancer¹², colon cancer¹³, bladder cancer¹⁴, and gastric cancer¹⁵, was also reported, with evidence manifesting that CD44 could promote tumorigenesis and has the potential to be a molecular target in cancer therapy¹⁶. Notwithstanding, most studies on CD44 were restricted to specific cancer types, obscuring its exact role in tumorigenesis. Thus, it is imperative to deeply explore the role of CD44 using pan-cancer analysis, thereby providing a novel avenue for developing novel treatments and individualized therapies.

The present study systematically assessed the predictive significance of CD44 expression in 33 cancer types using The Cancer Genome Atlas (TCGA) database. Then, the potential correlation of CD44 with tumour mutational burden (TMB) and microsatellite instability (MSI) was evaluated. Moreover, the underlying mechanisms of CD44 were also examined using the ESTIMATE and CIBERSORT algorithms and Gene Set Enrichment Analysis (GSEA). Our pan-cancer investigation aimed to illustrate the association of CD44 with immunotherapy response in oncogenesis among varying cancer types and its potential for predicting prognosis in pan-cancer patients.

CD44 expression analysis in pan-cancer

CD44 expression information, including RNA sequences, somatic mutations, and related clinical annotations for 33 cancers (Table 1), were downloaded from the University of California Santa Cruz (UCSC) Xena database.

Table 1

The abbreviation of 33 cancer types from TCGA database
Cancer types	Abbreviation
adrenocortical carcinoma	ACC
bladder urothelial carcinoma	BLCA
breast invasive carcinoma	BRCA
cervical squamous cell carcinoma and endocervical adenocarcinoma	CESC
cholangiocarcinoma	CHOL
colon adenocarcinoma	COAD
lymphoid neoplasm diffuse large B-cell lymphoma	DLBC
esophageal carcinoma	ESCA
glioblastoma multiforme	GBM
head and neck squamous cell carcinoma	HNSC
kidney chromophobe	KICH
kidney renal clear cell carcinoma	KIRC
kidney renal papillary cell carcinoma	KIRP
acute myeloid leukemia	LAML
brain lower grade glioma	LGG
liver hepatocellular carcinoma	LIHC
lung adenocarcinoma	LUAD
lung squamous cell carcinoma	LUSC
mesothelioma	MESO
ovarian serous cystadenocarcinoma	OV
pancreatic adenocarcinoma	PAAD
pheochromocytoma and paraganglioma	PCPG
prostate adenocarcinoma	PRAD
rectum adenocarcinoma	READ
sarcoma	SARC
skin cutaneous melanoma	SKCM
stomach adenocarcinoma	STAD
testicular germ cell tumors	TGCT
thyroid carcinoma	THCA
thymoma	THYM
uterine corpus endometrial carcinoma	UCEC
uterine carcinosarcoma	UCS
uveal melanoma	UVM

Afterward, CD44 expression was analyzed through the Tumor IMmune Estimation Resource (TIMER) database. The gene expression levels were normalized using log2 conversion. Meanwhile, R-package “ggpubr” was employed to visualize the results obtained from the TCGA database.

Association analyses between CD44 and survival data

Survival information for each sample was retrieved from the TCGA database, and the correlation of CD44 expression with survival indicators, including overall survival (OS), progression-free interval (PFI), disease-free interval (DFI), and disease-specific survival (DSS), were examined by the Kaplan-Meier method. Patients of varying cancer types were divided into two groups (high-risk and low-risk groups) using the median CD44 expression level as the cut-off value. The above analyses were performed using the R packages “survminer” and “survival”. In addition, univariate Cox analysis was used to score the correlation between CD44 expression and the survival indicators using the R package “survival”, and visualized by the R package “forestplot”.

Biomarker analysis of CD44 as an indicator of response to targeted therapy

ROC plotter database^17,18 was used to assess the feasibility of CD44 in targeted therapy responses to various cancer cohorts such as breast carcinoma (BRCA) treated with anti-HER2 antibody, colorectal carcinoma treated with bevacizumab, and ovarian cancer cohorts treated with targeted therapy. The receiver operating characteristic (ROC) curves was used to present predictive ability of CD44 in predicting efficacy of various antitumor agents in the above analysis.

Association between the expression of CD44 and clinical annotations, TMB and MSI

Clinical annotations (age, gender, and tumor stage) of pan-cancer patients were downloaded from the TCGA database, and Spearman correlation analysis between CD44 expression and clinical annotations was performed by the R packages “limma” and “ggpubr”. The correlation of CD44 expression with MSI or TMB was analyzed using Spearman analysis and visualized by radar plots through the R package “fmsb”. Next, Tumor Immune Dysfunction and Exclusion (TIDE) database was used to assess the potential of CD44 as a responsive biomarker for the cancer cohorts treated with immunotherapy (http://tide.dfci.harvard.edu/), which is a web application integrating the expression profiles of T cell dysfunction and exclusion, thereby modeling immune evasion of tumor cells, and has the potential to predict response of immune checkpoint blockade (ICB)¹⁹.

Relationship between CD44 expression, immune components, and tumor-infiltrating immune cell profiles

To determine the proportion of immune and stromal components in the TME, the ESTIMATE algorithm was applied to evaluate the associations between the immune and stromal scores with CD44 expression levels through R packages “estimate” and “limma”. Next, relative tumor-infiltrating immune cells (TICs) levels were calculated using the CIBERSORT algorithm, and samples of the tumor with P < 0.001 were retained for subsequent evaluations. Correlation analysis between CD44 expression and relative TICs levels was conducted using the R packages “ggplot2”, “ggpubr”, and “ggExtra”.

Immune-related genes and enrichment analysis in various cancer types

The R package “limma” was employed to perform co-expression analysis between CD44 and immune-related genes, and the results were visualized by “reshape2” and “RColorBrewer” R packages. Then, GSEA analysis was carried out to explore the role of CD44 in pan-cancers, and the top five enrichment terms of each tumor type were illustrated using the R package “ClusterProfiler”.

Statistical analysis

Alterations of CD44 expression in tumor tissues and normal tissues were estimated using the Wilcoxon test. In survival analysis, the relationship between CD44 expression and survival information in pan-cancer patients was determined using Kaplan-Meier and univariate Cox regression analyses. The evaluation of these data was conducted through R software (Version 4.0.3) and Strawberry Perl (Version 5.30.0.1). P < 0.05 was considered statistically significant.

Pan-Cancer Expression Profiles of CD44

To analyze the expression profiles of CD44 in the pan-cancer dataset, a comparative analysis of CD44 expression was performed between cancer and control samples using the TIMER database. And the results revealed that CD44 was up-regulated in cancer tissues matched with adjacent normal tissues in CHOL, COAD, ESCA, HNSC, KICH, KIRC, KIRP, READ, STAD, and THCA. In contrast, it was down-regulated in BLCA, LUAD, PRAD, and UCEC (Fig. 1A). Meanwhile, in the TCGA database, the CD44 levels were significantly up-regulated in CHOL, COAD, ESCA, GBM, HNSC, KICH, KIRC, KIRP, PCPG, READ, STAD, and THCA, whereas it was downregulated in LUAD, PRAD, and UCEC (Fig. 1B). CD44 expression levels were significantly up-regulated in CHOL, COAD, ESCA, HNSC, KICH, KIRC, KIRP, READ, STAD, and THCA, while downregulated in LUAD, PRAD, and UCEC in both databases. Intriguingly, we discovered that the results obtained by TIMER and TCGA databases varied, which may be due to the differences in sample size or data processing. Collectively, these results revealed the difference in CD44 expression patterns between cancer and normal samples in the pan-cancer datasets.

Correlation analysis between CD44 expression and survival

Then, correlation analysis of CD44 and the prognosis of pan-cancer patients were conducted. Survival indicators included OS, PFI, DFI, and DSS. In the OS analysis, the Kaplan–Meier survival curves indicated that the high expression of CD44 was remarkably associated with poor OS in KIRC (P = 0.013), LGG (P = 0.001), MESO (P = 0.002), PAAD (P = 0.008). On the other hand, subjects with lower CD44 expression were associated with poor OS in LAML (P = 0.015), UVM (P = 0.007) (Fig. 2A). In the PFI analysis, Kaplan–Meier analysis showed that patients with higher CD44 expression had a shorter PFI in GBM (P = 0.027), LGG (P < 0.001), THYM (P = 0.025), whereas those with higher CD44 expression showed a better PFI in BRCA (P = 0.046), PCPG (P = 0.041), PRAD (P = 0.037), UCEC (P = 0.032), UVM (P = 0.044) (Fig. 2B). Likewise, results of Kaplan–Meier analysis indicated that patients with higher CD44 expression had a poorer DFI in PAAD (P = 0.004), whereas patients with higher CD44 levels showed a better DFI in CESC (P = 0.024) (Fig. 2C). Besides, Kaplan–Meier analysis indicated that the increased CD44 expression correlated with poorer DSS in patients with BLCA (P = 0.036), KIRC (P = 0.006), LGG (P < 0.001), PAAD (P = 0.019). Subjects with decreased CD44 expression manifested poorer DSS in patients with BRCA (P = 0.027), UCEC (P = 0.007), UVM (P = 0.007) (Fig. 2D).

Cox regression of OS identified that CD44 expression was a risk factor for KIRC (P < 0.001), KIRP (P = 0.027), LGG (P < 0.001), MESO (P = 0.025), PAAD (P < 0.001), however, it appeared to be a protective factor in SARC (P = 0.019), UVM (P = 0.002) (Fig. 3A). Cox regression analysis of DSS demonstrated that CD44 expression as a risk factor in BLCA (P = 0.043), HNSC (P = 0.033), KIRC (P < 0.001), KIRP (P = 0.023), LGG (P < 0.001), PAAD (P = 0.004), while it was a protective factor in BRCA (P < 0.001), SARC (P = 0.012), UCEC (P = 0.033), UVM (P = 0.002) (Fig. 3B). Cox regression analysis of DFI noted that CD44 expression was a risk factor in PAAD (P = 0.001), while it was a protective factor in BRCA (P = 0.038), OV (P = 0.038) (Fig. 3C). Cox regression analysis of PFI identified that CD44 acted as a detrimental prognostic factor in GBM (P = 0.010), HNSC (P = 0.036), KIRC (P < 0.001), LGG (P < 0.001), PAAD (P < 0.001), THYM (P = 0.037), while it was a protective prognostic factor in BRCA (P = 0.013), PCPG (P = 0.024), PRAD (P = 0.041), UCEC (P = 0.020), UVM (P = 0.009) (Fig. 3D) (Table 2). Altogether, these results signal that CD44 may serve as a prognostic biomarker and potential therapeutic target.

Table 2

Univariate Cox regression analysis of the associations of CD44 expression with patient survival
Cancer	OS			PFI			DFI			DSS
Cancer	HR	HR(95%CI)	P-value	HR	HR(95%CI)	P-value	HR	HR(95%CI)	P-value	HR	HR(95%CI)	P-value
ACC	0.964	(0.705 − 1.320)	0.821	0.874	(0.675 − 1.130)	0.303	0.977	(0.605 − 1.579)	0.925	0.889	(0.635 − 1.244)	0.492
BLCA	1.065	(0.966 − 1.173)	0.205	1.044	(0.945 − 1.153)	0.401	0.955	(0.756 − 1.206)	0.697	1.134	(1.004 − 1.281)	0.043
BRCA	0.924	(0.788 − 1.082)	0.324	0.817	(0.697 − 0.959)	0.013	0.798	(0.645 − 0.988)	0.038	0.700	(0.568 − 0.864)	< 0.001
CESC	1.069	(0.857 − 1.333)	0.555	1.005	(0.807 − 1.251)	0.965	0.782	(0.547 − 1.118)	0.177	1.012	(0.788 − 1.300)	0.924
CHOL	0.907	(0.668 − 1.232)	0.532	0.794	(0.598 − 1.055)	0.112	1.060	(0.731 − 1.538)	0.759	0.903	(0.652 − 1.249)	0.537
COAD	0.993	(0.736 − 1.339)	0.961	0.929	(0.701 − 1.232)	0.610	0.859	(0.467 − 1.577)	0.623	0.984	(0.673 − 1.439)	0.934
DLBC	0.800	(0.359 − 1.779)	0.584	0.795	(0.396 − 1.596)	0.519	0.950	(0.174 − 5.200)	0.953	1.068	(0.357 − 3.200)	0.906
ESCA	0.969	(0.765 − 1.226)	0.791	1.080	(0.876 − 1.331)	0.470	1.363	(0.912 − 2.039)	0.131	1.019	(0.772 − 1.344)	0.895
GBM	1.169	(0.990 − 1.380)	0.065	1.253	(1.054 − 1.489)	0.010	/	/	/	1.157	(0.965 − 1.386)	0.114
HNSC	1.176	(0.977 − 1.415)	0.086	1.236	(1.014 − 1.508)	0.036	1.191	(0.681 − 2.083)	0.539	1.321	(1.023 − 1.707)	0.033
KICH	1.051	(0.638 − 1.732)	0.844	1.050	(0.668 − 1.652)	0.831	0.772	(0.340 − 1.751)	0.535	1.136	(0.640 − 2.019)	0.663
KIRC	1.476	(1.269 − 1.718)	< 0.001	1.384	(1.182 − 1.620)	< 0.001	0.827	(0.485 − 1.411)	0.486	1.690	(1.408 − 2.030)	< 0.001
KIRP	1.289	(1.029 − 1.615)	0.027	1.112	(0.905 − 1.367)	0.312	1.073	(0.785 − 1.467)	0.657	1.363	(1.044 − 1.779)	0.023
LAML	0.779	(0.595 − 1.019)	0.068	/	/	/	/	/	/	/	/	/
LGG	1.416	(1.232 − 1.627)	< 0.001	1.318	(1.178 − 1.473)	< 0.001	1.223	(0.881 − 1.697)	0.229	1.481	(1.273 − 1.723)	< 0.001
LIHC	1.131	(0.984 − 1.300)	0.083	1.006	(0.891 − 1.136)	0.924	0.960	(0.834 − 1.105)	0.572	1.094	(0.908 − 1.319)	0.345
LUAD	1.021	(0.881 − 1.183)	0.782	1.004	(0.872 − 1.157)	0.952	1.011	(0.817 − 1.250)	0.923	1.006	(0.830 − 1.219)	0.953
LUSC	0.928	(0.796 − 1.083)	0.345	0.938	(0.778 − 1.133)	0.508	0.947	(0.706 − 1.269)	0.714	0.822	(0.653 − 1.034)	0.094
MESO	1.305	(1.033 − 1.649)	0.025	1.094	(0.863 − 1.386)	0.457	1.286	(0.700 − 2.361)	0.417	1.227	(0.919 − 1.639)	0.166
OV	0.895	(0.777 − 1.031)	0.125	0.882	(0.776 − 1.002)	0.054	0.817	(0.675 − 0.989)	0.038	0.869	(0.745 − 1.013)	0.073
PAAD	1.650	(1.245 − 2.188)	< 0.001	1.665	(1.257 − 2.206)	< 0.001	2.882	(1.508 − 5.505)	0.001	1.574	(1.154 − 2.147)	0.004
PCPG	0.780	(0.495 − 1.227)	0.282	0.723	(0.546 − 0.958)	0.024	0.716	(0.365 − 1.404)	0.331	0.742	(0.449 − 1.227)	0.245
PRAD	0.890	(0.504 − 1.572)	0.688	0.805	(0.655 − 0.991)	0.041	0.837	(0.575 − 1.218)	0.353	0.592	(0.272 − 1.285)	0.185
READ	0.628	(0.334 − 1.179)	0.147	0.843	(0.477 − 1.492)	0.559	0.682	(0.107 − 4.347)	0.685	1.047	(0.387 − 2.832)	0.928
SARC	0.828	(0.707 − 0.970)	0.019	0.943	(0.823 − 1.080)	0.394	0.961	(0.784 − 1.178)	0.702	0.801	(0.674 − 0.952)	0.012
SKCM	0.936	(0.827 − 1.059)	0.295	0.939	(0.844 − 1.045)	0.252	/	/	/	0.938	(0.820 − 1.073)	0.353
STAD	1.129	(0.978 − 1.304)	0.097	1.157	(0.992 − 1.348)	0.063	1.109	(0.836 − 1.472)	0.474	1.181	(0.986 − 1.413)	0.070
TGCT	2.256	(0.922 − 5.523)	0.075	0.926	(0.679 − 1.262)	0.626	0.871	(0.605 − 1.253)	0.457	2.226	(0.901 − 5.501)	0.083
THCA	0.724	(0.386 − 1.358)	0.314	1.122	(0.767 − 1.641)	0.552	1.343	(0.785 − 2.295)	0.282	0.492	(0.217 − 1.114)	0.089
THYM	1.674	(0.814 − 3.441)	0.161	1.697	(1.033 − 2.788)	0.037	/	/	/	1.157	(0.381 − 3.519)	0.797
UCEC	0.887	(0.740 − 1.062)	0.193	0.831	(0.711 − 0.972)	0.020	0.830	(0.662 − 1.042)	0.108	0.785	(0.628 − 0.981)	0.033
UCS	0.903	(0.646 − 1.262)	0.55	0.950	(0.701 − 1.289)	0.742	0.923	(0.500 − 1.705)	0.798	0.834	(0.578 − 1.203)	0.331
UVM	0.305	(0.143 − 0.650)	0.002	0.392	(0.194 − 0.794)	0.009	/	/	/	0.294	(0.136 − 0.637)	0.002
Abbreviations: OS, overall survival; PFI, progression-free interval; DFI, disease-free interval; DSS, disease-specific survival.

Correlation analysis between CD44 expression and pan-cancer clinicopathologic characteristics

Thereafter, the association between CD44 expression and clinicopathologic characteristics was investigated in pan-cancer datasets. In patients less than (or equal to) 65 years, a higher CD44 expression level was noted in ESCA, THYM, and UCEC. In contrast, CD44 was higher expressed in patients over 65 years old in LUAD (Fig. 4A). Moreover, CD44 was down-regulated during the advanced cancer clinical stage in BRCA, COAD, and SKCM. Notably, CD44 was increased as the clinical stage changed from Ⅰ&Ⅱ to Ⅲ&Ⅳ in KIRC (Fig. 4B). Interestingly, CD44 expression also presented sex dimorphism in KIRC, LIHC, and LUAD (Fig. 4C), implying that CD44 may reflect clinical progression for those tumors.

Correlation analysis of CD44 expression with TMB and MSI

The TMB and MSI status in CD44 expression was further evaluated to determine the potential of CD44 in reflecting the efficacy of immunotherapy. High TMB was reported as a critical driver of cancer progression²⁰. Our results determined a positive correlation between CD44 expression and TMB in COAD, LGG, and UCEC. In contrast, CD44 expression was negatively linked with TMB in BLCA, BRCA, ESCA, LIHC, LUAD, LUSC, PRAD, and TGCT (Fig. 5A).

Meanwhile, MSI also acted as a predictive biomarker, enabling more precise guidance of immunotherapy²¹. Hence, the relationship of CD44 expression with MSI was analyzed. Our findings revealed a positive correlation between CD44 and MSI in COAD and UCEC. On the other hand, a negative correlation was discovered between CD44 and MSI in DLBC, ESCA, HNSC, KIRC, LUAD, PRAD, and TGCT (Fig. 5B). Next, we compared CD44’s predictive ability for immunotherapy efficacy to other canonical biomarker signatures in TIDE database, using treatment response from various cancer cohorts treated with ICB. The results confirmed that CD44 had the medium predictive performance, with 10 of the 25 ICB-treated cohorts presenting an area under curve (AUC) greater than 0.5 (Fig. 6)

To sum up, these observations indicated that TMB and MSI were correlated in multiple cancer types, and the results provided by TIDE database persuasively confirm its robustness in efficacy predictions, undoubtfully manifesting that it could be used as a reliable biomarker for predicting responses to immunotherapy.

Associations between CD44 expression and therapeutic response of targeted therapy in various cancer

We investigated the utility of CD44 in evaluating therapeutic responses of targeted therapy to various cancer (Fig. 7). In BRCA treated with anti-HER2 therapy, CD44 expression level was higher in non-responders, and with an AUC of 0.588. Likewise, CD44 expression was higher in non-responders in colorectal carcinoma treated with bevacizumab, with an AUC of 0.64. Furthermore, CD44 was associated with benefits of targeted therapeutic relapse-free survival (RFS) at 12 months in the ovarian cancer, with an AUC of 0.733. Taken together, above results elucidated that CD44 could act as a therapeutic response biomarker in various cancer.

Correlation between CD44 expression and various components in the TME of pan-cancer

Components of the TME include tumor cells, stromal cells, and immune cells²², which can influence tumor formation, maintenance, and multidrug resistance, whereas non-malignant cells promote tumorigenesis in all stages of cancer²³. Subsequently, the pan-cancer types were divided into high-risk cancer groups (BLCA, KIRC, KIRP, and LGG) (Fig. 8A) and low-risk cancer groups (OV, PRAD, SARC, and UCEC) (Fig. 8B) according to the survival outcome of the pan-cancer patients obtained from Kaplan–Meier curve and univariate Cox analysis. A positive correlation was identified between CD44 expression and the two cancer groups.

To further explore the association of CD44 expression and TIC subtypes, the CIBERSORT algorithm was utilized to calculate the relative levels of TIC subtypes in patients from both groups, with P < 0.001 as the cut-off value. It was observed that CD44 expression levels were positively linked with neutrophils in BLCA and negatively linked with naive B cells, plasma cells, and regulatory T cells (Tregs) (Fig. 9A-D). Moreover, CD44 expression was positively correlated with macrophages M0, activated memory CD4 T cells, and Tregs in KIRC, and negatively correlated with resting mast cells, monocytes, resting NK cells, and resting memory CD4 T cells (Fig. 9E-K). In addition, CD44 expression levels were significantly positively correlated with naive B cells, neutrophils, and activated memory CD4 T cells in KIRP but negatively correlated with memory B cells and macrophages M2 (Fig. 9L-P). Besides, a positive correlation between CD44 expression and resting memory CD4 T cells was noted in LGG (Fig. 9Q). In OV, CD44 expression levels were positively correlated with resting dendritic cells, neutrophils, and plasma cells and negatively correlated with activated dendritic cells (Fig. 9R-U). In UCEC, CD44 expression levels were positively correlated with neutrophils and activated memory CD4 T cells and negatively correlated with activated NK cells and memory B cells (Fig. 9V-Y). Collectively, these results infer that CD44 may mediate the immune response in these cancer types. Thereupon, correlation analysis was performed between CD44 and various immune-related genes (Fig. 10A), and immune checkpoint genes (Fig. 10B). Finally, the results suggested that CD44 may interfere with the TME by influencing the expression of immune-related genes and immune checkpoint genes, which mediates tumor progression and metastasis.

Cancers enrichment analysis

To elucidate the underlying molecular mechanism of the CD44 in tumorigenesis, GSEA was performed to assess the biological significance of CD44 expression in eight pan-cancer types (Fig. 11). In Gene Ontology (GO) functional annotation, CD44 was significantly correlated with several immune-related functions in KIRC, KIRP, and UCEC, such as leukocyte migration and detection of chemical stimulus (Fig. 11A). Furthermore, Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis demonstrated that CD44 could positively influence several crucial immune cell-related pathways in KIRP and LGG, such as toll-like receptor signaling pathway and Leishmania infection (Fig. 11B). Overall, these results confirmed that CD44 is instrumental in TME remodeling for various cancers.

The pan-cancer analysis can disclose the heterogeneities of tumors, providing insights into cancer treatment²⁴. Numerous pan-cancer studies have focused on gene mutations and cancer development, which are helpful in the progression of sustainable, meaningful clinical treatments and the development of biomarkers²⁵. As previously reported, CD44 is overexpressed in CSCs and plays a vital role in cancer progression, metastasis, and drug resistance. It may also serve as a therapeutic target, given that it modulates multiple survival signaling pathways^9,26. It has been previously reported that CD44 can promote CSCs traits of metastatic breast cancers by activating the PDGFRβ/Stat3 signaling pathway¹⁰. Accumulating evidence revealed that CD44 might present as a therapeutic biomarker in various tumor types^11,27. Although the CD44 has been extensively studied in certain types of cancer, its role remains elusive in multiple cancers. In this research, we described the functional significance of CD44 and identified the differential expression of CD44 within cancers and normal tissues in 15 cancer types from the pan-cancer datasets. Moreover, we also confirmed that CD44 expression was relevant to the levels of immune cell infiltration in various types of cancer with ESTIMATE and CIBERSORT. Lastly, GSEA analysis exposed that CD44 was significantly correlated with several signaling pathways.

CSCs are hypothesized to possess the ability of self-renewal, tumor initiation and metastasis. Prior research reported that the overexpression of CD44 in cancer cells is widely accepted as a marker of higher tumor-initiating potential and invasiveness of cancer cells^28,29. CD44 is recognized as the CSC surface marker for sorting cancer types such as breast cancer⁸, prostate cancer³⁰, and gastric cancers¹⁵. Previous studies revealed that CD44 might be unnaturally expressed in several cancer types and plays an essential role in cancer progression. Herein, significant upregulation of CD44 expression levels was observed in cancer tissues compared to normal tissues in CHOL, COAD, ESCA, HNSC, KICH, KIRC, KIRP, READ, STAD, and THCA, while down-regulated in LUAD, PRAD, and UCEC. Interestingly, some studies reported contrasting outcomes. For instance, a study reported that CD44 was up-regulated in LUAD, showed significantly higher capacities of tumorigenic colonies³¹, and was related to worse OS³². Notably, the functional role of CD44 on cancer development and progression has become a research hotspot and will assist in understanding its potential role as a prognostic biomarker for cancers.

Furthermore, our results established that a higher CD44 expression level was related to unfavorable survival outcomes in BLCA, KIRC, GBM, LGG, MESO, PAAD, and THYM. Similar outcomes were also observed in glioma patients³³. Compelling evidence obtained from 42 studies outlined that gastric cancer patients with CD44 overexpression had a lower 5-year OS rate³⁴. Some studies also reported similar results in colorectal cancer^35,36. Moreover, overexpression of CD44 predicted a poor prognosis in patients with hepatocellular carcinoma³⁷ and pancreatic carcinoma³⁸. Besides, another study revealed that expression of CD44 varied significantly by age and gender in oral cancer³⁹, which is consistent with the outcomes of this study, where CD44 was up-regulated in older patients with LUAD and downregulated in older patients affected by ESCA, THYM, and UCEC. In short, these outcomes strongly indicate that CD44 might be a useful biomarker for most cancer types.

Gene mutation is postulated to be the primary cause of cancer⁴⁰, and specific gene mutations have distinct impacts on the prognosis and risk stratification of various cancer types⁴¹. TMB is defined as the number of somatic mutations per megabase of the interrogated genomic sequence, while MSI is defined as the collection of microsatellites mutations; both are widely used as predictive biomarkers of response to immunotherapy^42,43. Additionally, recent studies have demonstrated that MSI and TMB contribute significantly to the therapeutic response to immune checkpoint inhibitors (ICIs)⁴⁴. The MSI-low phenotype was found as a worse prognostic biomarker in colorectal cancers⁴⁵. However, MSI has limitations, such as immune checkpoint blockade failing to elicit a response in colorectal cancer cases⁴⁶. This research established a relationship between CD44 and TMB and MSI, implying that CD44 may provide a more comprehensive perspective of immunotherapy in these cancer types.

The MSI status may alter the TME of cancer patients, thereby affecting the efficacy of ICIs⁴⁷, while TME plays a crucial role in tumorigenesis and cancer progression^48,49. Increasing evidence indicated that the immune escape of cancer cells is correlated with various components of the TME and ultimately contributes to tumor proliferation, metastasis, and recurrence. Albeit immunotherapy has made considerable advances in cancer treatment, it still faces numerous challenges in its successful application^50,51. Indeed, to further improve the efficacy of immunotherapy, the identification of novel biomarkers is vital. Gomez et al. reported that CD44 expression was regulated by TAM, which directly influences CD44 signaling via ligand binding in HNSC¹¹. Nonetheless, little is known about the role of CD44 in the immune microenvironment. The results from this study indicated that CD44 level was significantly correlated with T cells, B cells, NK cells, macrophages, and other immune infiltrating cells in BLCA, KIRC, KIRP, LGG, OV, and UCEC. Taken together, it is reasonable to speculate that CD44 may play an essential role in cancer immunity and ultimately influence prognosis. This study still has some limitation, biological validation and large sample cancer cohort validation should be performed to better illustrate the role of CD44 in pan-cancer study.

In summary, our results indicated that CD44 was associated with disease prognosis and immune infiltration in pan-cancers. Moreover, CD44 expression was also linked with TMB, MSI, and various components of the TME. These findings add to the understanding of tumor mechanisms and contribute to improving the efficacy of immunotherapy.

Data availability statement

The datasets supporting the conclusions of this article are available in the the University of California Santa Cruz Xena (UCSC) Xena repository (http://xena.ucsc.edu/).

Ethical statement

Our study did not require an ethical board approval because it did not contain human or animal trials.

Author Contributions

Shaoyong Chen and Siqin Zhang wrote the main manuscript text and data analysis; Shaohua Chen performed the study design and reviewed the research; Fei Ma reviewed and supervised the research. All authors reviewed the final version of the manuscript.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Conflict of Interests

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Sung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA: a cancer journal for clinicians. May 2021;71(3):209-249. doi:10.3322/caac.21660
Fitzmaurice C, Abate D, Abbasi N, et al. Global, Regional, and National Cancer Incidence, Mortality, Years of Life Lost, Years Lived With Disability, and Disability-Adjusted Life-Years for 29 Cancer Groups, 1990 to 2017: A Systematic Analysis for the Global Burden of Disease Study. JAMA oncology. Dec 1 2019;5(12):1749-1768. doi:10.1001/jamaoncol.2019.2996
Siegel RL, Miller KD, Jemal A. Cancer statistics, 2019. CA: a cancer journal for clinicians. Jan 2019;69(1):7-34. doi:10.3322/caac.21551
Li B, Severson E, Pignon JC, et al. Comprehensive analyses of tumor immunity: implications for cancer immunotherapy. Genome biology. Aug 22 2016;17(1):174. doi:10.1186/s13059-016-1028-7
Kelderman S, Schumacher TN, Haanen JB. Acquired and intrinsic resistance in cancer immunotherapy. Molecular oncology. Sep 12 2014;8(6):1132-9. doi:10.1016/j.molonc.2014.07.011
Liu B, Fan Y, Song Z, et al. Identification of DRP1 as a prognostic factor correlated with immune infiltration in breast cancer. International immunopharmacology. Dec 2020;89(Pt B):107078. doi:10.1016/j.intimp.2020.107078
Russnes HG, Lønning PE, Børresen-Dale AL, Lingjærde OC. The multitude of molecular analyses in cancer: the opening of Pandora’s box. Genome biology. 2014;15(9):447. doi:10.1186/s13059-014-0447-6
Al-Hajj M, Wicha MS, Benito-Hernandez A, Morrison SJ, Clarke MF. Prospective identification of tumorigenic breast cancer cells. Proceedings of the National Academy of Sciences of the United States of America. Apr 1 2003;100(7):3983-8. doi:10.1073/pnas.0530291100
Mattheolabakis G, Milane L, Singh A, Amiji MM. Hyaluronic acid targeting of CD44 for cancer therapy: from receptor biology to nanomedicine. Journal of drug targeting. 2015;23(7-8):605-18. doi:10.3109/1061186x.2015.1052072
Zhang H, Brown RL, Wei Y, et al. CD44 splice isoform switching determines breast cancer stem cell state. Genes & development. Feb 1 2019;33(3-4):166-179. doi:10.1101/gad.319889.118
Gomez KE, Wu F, Keysar SB, et al. Cancer Cell CD44 Mediates Macrophage/Monocyte-Driven Regulation of Head and Neck Cancer Stem Cells. Cancer research. Oct 1 2020;80(19):4185-4198. doi:10.1158/0008-5472.Can-20-1079
Kalantari E, Asgari M, Nikpanah S, Salarieh N, Asadi Lari MH, Madjd Z. Co-Expression of Putative Cancer Stem Cell Markers CD44 and CD133 in Prostate Carcinomas. Pathology oncology research : POR. Oct 2017;23(4):793-802. doi:10.1007/s12253-016-0169-z
Todaro M, Gaggianesi M, Catalano V, et al. CD44v6 is a marker of constitutive and reprogrammed cancer stem cells driving colon cancer metastasis. Cell stem cell. Mar 6 2014;14(3):342-56. doi:10.1016/j.stem.2014.01.009
Kobayashi K, Matsumoto H, Matsuyama H, et al. Clinical significance of CD44 variant 9 expression as a prognostic indicator in bladder cancer. Oncology reports. Nov 2016;36(5):2852-2860. doi:10.3892/or.2016.5061
Lau WM, Teng E, Chong HS, et al. CD44v8-10 is a cancer-specific marker for gastric cancer stem cells. Cancer research. May 1 2014;74(9):2630-41. doi:10.1158/0008-5472.Can-13-2309
Li L, Hao X, Qin J, et al. Antibody against CD44s inhibits pancreatic tumor initiation and postradiation recurrence in mice. Gastroenterology. Apr 2014;146(4):1108-18. doi:10.1053/j.gastro.2013.12.035
Fekete JT, Győrffy B. ROCplot.org: Validating predictive biomarkers of chemotherapy/hormonal therapy/anti-HER2 therapy using transcriptomic data of 3,104 breast cancer patients. International journal of cancer. Dec 1 2019;145(11):3140-3151. doi:10.1002/ijc.32369
Fekete JT, Ősz Á, Pete I, Nagy GR, Vereczkey I, Győrffy B. Predictive biomarkers of platinum and taxane resistance using the transcriptomic data of 1816 ovarian cancer patients. Gynecologic oncology. Mar 2020;156(3):654-661. doi:10.1016/j.ygyno.2020.01.006
Jiang P, Gu S, Pan D, et al. Signatures of T cell dysfunction and exclusion predict cancer immunotherapy response. Nature medicine. Oct 2018;24(10):1550-1558. doi:10.1038/s41591-018-0136-1
Chalmers ZR, Connelly CF, Fabrizio D, et al. Analysis of 100,000 human cancer genomes reveals the landscape of tumor mutational burden. Genome medicine. Apr 19 2017;9(1):34. doi:10.1186/s13073-017-0424-2
Dudley JC, Lin MT, Le DT, Eshleman JR. Microsatellite Instability as a Biomarker for PD-1 Blockade. Clinical cancer research : an official journal of the American Association for Cancer Research. Feb 15 2016;22(4):813-20. doi:10.1158/1078-0432.Ccr-15-1678
Jahanban-Esfahlan R, Seidi K, Zarghami N. Tumor vascular infarction: prospects and challenges. International journal of hematology. Mar 2017;105(3):244-256. doi:10.1007/s12185-016-2171-3
Frisch J, Angenendt A, Hoth M, Prates Roma L, Lis A. STIM-Orai Channels and Reactive Oxygen Species in the Tumor Microenvironment. Cancers. Mar 30 2019;11(4)doi:10.3390/cancers11040457
Schaub FX, Dhankani V, Berger AC, et al. Pan-cancer Alterations of the MYC Oncogene and Its Proximal Network across the Cancer Genome Atlas. Cell systems. Mar 28 2018;6(3):282-300.e2. doi:10.1016/j.cels.2018.03.003
Consortium ITP-CAoWG. Pan-cancer analysis of whole genomes. Nature. Feb 2020;578(7793):82-93. doi:10.1038/s41586-020-1969-6
Chen C, Zhao S, Karnad A, Freeman JW. The biology and role of CD44 in cancer progression: therapeutic implications. Journal of hematology & oncology. May 10 2018;11(1):64. doi:10.1186/s13045-018-0605-5
Heldin P, Kolliopoulos C, Lin CY, Heldin CH. Involvement of hyaluronan and CD44 in cancer and viral infections. Cellular signalling. Jan 2020;65:109427. doi:10.1016/j.cellsig.2019.109427
Chaffer CL, Brueckmann I, Scheel C, et al. Normal and neoplastic nonstem cells can spontaneously convert to a stem-like state. Proceedings of the National Academy of Sciences of the United States of America. May 10 2011;108(19):7950-5. doi:10.1073/pnas.1102454108
La Fleur L, Johansson AC, Roberg K. A CD44high/EGFRlow subpopulation within head and neck cancer cell lines shows an epithelial-mesenchymal transition phenotype and resistance to treatment. PloS one. 2012;7(9):e44071. doi:10.1371/journal.pone.0044071
Collins AT, Berry PA, Hyde C, Stower MJ, Maitland NJ. Prospective identification of tumorigenic prostate cancer stem cells. Cancer research. Dec 1 2005;65(23):10946-51. doi:10.1158/0008-5472.Can-05-2018
Wang CY, Huang CS, Yang YP, et al. The subpopulation of CD44-positive cells promoted tumorigenicity and metastatic ability in lung adenocarcinoma. Journal of the Chinese Medical Association : JCMA. Mar 2019;82(3):196-201. doi:10.1097/jcma.0000000000000056
Zhang C, Wang H, Wang X, Zhao C, Wang H. CD44, a marker of cancer stem cells, is positively correlated with PD-L1 expression and immune cells infiltration in lung adenocarcinoma. Cancer cell international. Dec 7 2020;20(1):583. doi:10.1186/s12935-020-01671-4
Wu G, Song X, Liu J, et al. Expression of CD44 and the survival in glioma: a meta-analysis. Bioscience reports. Apr 30 2020;40(4)doi:10.1042/bsr20200520
Fang M, Wu J, Lai X, et al. CD44 and CD44v6 are Correlated with Gastric Cancer Progression and Poor Patient Prognosis: Evidence from 42 Studies. Cellular physiology and biochemistry : international journal of experimental cellular physiology, biochemistry, and pharmacology. 2016;40(3-4):567-578. doi:10.1159/000452570
Wang JL, Su WY, Lin YW, et al. CD44v6 overexpression related to metastasis and poor prognosis of colorectal cancer: A meta-analysis. Oncotarget. Feb 21 2017;8(8):12866-12876. doi:10.18632/oncotarget.14163
Ma L, Dong L, Chang P. CD44v6 engages in colorectal cancer progression. Cell death & disease. Jan 10 2019;10(1):30. doi:10.1038/s41419-018-1265-7
Fu Y, Geng Y, Yang N, et al. CD44v6 expression is associated with a poor prognosis in Chinese hepatocellular carcinoma patients: A meta-analysis. Clinics and research in hepatology and gastroenterology. Dec 2015;39(6):736-9. doi:10.1016/j.clinre.2015.03.001
Chen K, Li Z, Jiang P, et al. Co-expression of CD133, CD44v6 and human tissue factor is associated with metastasis and poor prognosis in pancreatic carcinoma. Oncology reports. Aug 2014;32(2):755-63. doi:10.3892/or.2014.3245
Pereira LH, Reis IM, Reategui EP, et al. Risk Stratification System for Oral Cancer Screening. Cancer prevention research (Philadelphia, Pa). Jun 2016;9(6):445-55. doi:10.1158/1940-6207.Capr-15-0200
Sanz-Garcia E, Argiles G, Elez E, Tabernero J. BRAF mutant colorectal cancer: prognosis, treatment, and new perspectives. Annals of oncology : official journal of the European Society for Medical Oncology. Nov 1 2017;28(11):2648-2657. doi:10.1093/annonc/mdx401
Wu HX, Wang ZX, Zhao Q, et al. Tumor mutational and indel burden: a systematic pan-cancer evaluation as prognostic biomarkers. Annals of translational medicine. Nov 2019;7(22):640. doi:10.21037/atm.2019.10.116
Yarchoan M, Hopkins A, Jaffee EM. Tumor Mutational Burden and Response Rate to PD-1 Inhibition. The New England journal of medicine. Dec 21 2017;377(25):2500-2501. doi:10.1056/NEJMc1713444
Merino DM, McShane LM, Fabrizio D, et al. Establishing guidelines to harmonize tumor mutational burden (TMB): in silico assessment of variation in TMB quantification across diagnostic platforms: phase I of the Friends of Cancer Research TMB Harmonization Project. Journal for immunotherapy of cancer. Mar 2020;8(1)doi:10.1136/jitc-2019-000147
Eso Y, Shimizu T, Takeda H, Takai A, Marusawa H. Microsatellite instability and immune checkpoint inhibitors: toward precision medicine against gastrointestinal and hepatobiliary cancers. Journal of gastroenterology. Jan 2020;55(1):15-26. doi:10.1007/s00535-019-01620-7
Torshizi Esfahani A, Seyedna SY, Nazemalhosseini Mojarad E, Majd A, Asadzadeh Aghdaei H. MSI-L/EMAST is a predictive biomarker for metastasis in colorectal cancer patients. Journal of cellular physiology. Aug 2019;234(8):13128-13136. doi:10.1002/jcp.27983
Picard E, Verschoor CP, Ma GW, Pawelec G. Relationships Between Immune Landscapes, Genetic Subtypes and Responses to Immunotherapy in Colorectal Cancer. Frontiers in immunology. 2020;11:369. doi:10.3389/fimmu.2020.00369
Lin A, Zhang J, Luo P. Crosstalk Between the MSI Status and Tumor Microenvironment in Colorectal Cancer. Frontiers in immunology. 2020;11:2039. doi:10.3389/fimmu.2020.02039
Neal JT, Li X, Zhu J, et al. Organoid Modeling of the Tumor Immune Microenvironment. Cell. Dec 13 2018;175(7):1972-1988.e16. doi:10.1016/j.cell.2018.11.021
Chen S, Wang C, Su X, Dai X, Li S, Mo Z. KCNN4 is a potential prognostic marker and critical factor affecting the immune status of the tumor microenvironment in kidney renal clear cell carcinoma. Translational andrology and urology. Jun 2021;10(6):2454-2470. doi:10.21037/tau-21-332
Murciano-Goroff YR, Warner AB, Wolchok JD. The future of cancer immunotherapy: microenvironment-targeting combinations. Cell research. Jun 2020;30(6):507-519. doi:10.1038/s41422-020-0337-2
Chen S, Sun Y, Zhu X, Mo Z. Prediction of Survival Outcome in Lower-Grade Glioma Using a Prognostic Signature with 33 Immune-Related Gene Pairs. International journal of general medicine. 2021;14:8149-8160. doi:10.2147/ijgm.S338135

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

The Prognostic Value and Immunological Role of CD44 in Pan-Cancer Study

Status:

Version 1

Abstract

Background

Methods

Results

Conclusion

Figures

Introduction

Materials And Methods

CD44 expression analysis in pan-cancer

Association analyses between CD44 and survival data

Biomarker analysis of CD44 as an indicator of response to targeted therapy

Association between the expression of CD44 and clinical annotations, TMB and MSI

Relationship between CD44 expression, immune components, and tumor-infiltrating immune cell profiles

Immune-related genes and enrichment analysis in various cancer types

Statistical analysis

Results

Pan-Cancer Expression Profiles of CD44

Correlation analysis between CD44 expression and survival

Correlation analysis between CD44 expression and pan-cancer clinicopathologic characteristics

Correlation analysis of CD44 expression with TMB and MSI

Associations between CD44 expression and therapeutic response of targeted therapy in various cancer

Correlation between CD44 expression and various components in the TME of pan-cancer

Cancers enrichment analysis

Discussion

Conclusions

Declarations

References

Additional Declarations

Status:

Version 1