An exosome-derived lncRNA signature identified by machine learning associated with prognosis and biomarkers for immunotherapy in ovarian cancer

Abstract

Background: Ovarian cancer (OC) has the highest mortality rate among gynecological malignancies. Current treatment options are limited and ineffective, prompting the discovery of reliable biomarkers. Exosome lncRNAs, carrying genetic information, are promising new markers. Previous studies only focused on exosome-related genes and employed the Lasso algorithm to construct prediction models, which are not robust.

Methods: 420 OC patients from the TCGA datasets were divided into training and validation datasets. The GSE102037 dataset was used for external validation. LncRNAs associated with exosome-related genes were selected using Pearson analysis. Univariate COX regression analysis was used to filter prognosis-related lncRNAs. The overlapping lncRNAs were identified as candidate lncRNAs for machine learning. Based on 10 machine learning algorithms and 117 algorithm combinations, the optimal predictor combinations were selected according to the C index. The exosome-related LncRNA Signature (ERLS) model was constructed using multivariate COX regression. Based on the median risk score of the training datasets, the patients were divided into high- and low-risk groups. Kaplan-Meier survival analysis, the time-dependent ROC, immune cell infiltration, immunotherapy response, and immune checkpoints were analyzed.

Results: 64 lncRNAs were subjected to a machine-learning process. Based on the stepCox (forward) combined Ridge algorithm, 20 lncRNA were selected to construct the ERLS model. Kaplan-Meier survival analysis showed that the high-risk group had a lower survival rate. The area under the curve (AUC) in predicting OS at 1, 3, and 5 years were 0.758, 0.816, and 0.827 in the entire TCGA cohort. xCell and ssGSEA analysis showed that the low-risk group had higher immune cell infiltration, which may contribute to the activation of cytolytic activity, inflammation promotion, and T-cell co-stimulation pathways. The low-risk group had higher expression levels of PDL1, CTLA4, and higher TMB. The ERLS model can predict response to anti-PD1 and anti-CTLA4 therapy. Patients with low expression of PDL1 or high expression of CTLA4 and low ERLS exhibited significantly better survival prospects, whereas patients with high ERLS and low levels of PDL1 or CTLA4 exhibited the poorest outcomes.

Conclusion: Our study constructed an ERLS model that can predict prognostic risk and immunotherapy response, optimizing clinical management for OC patients.

Keywords: exosome-related lncRNA; immunotherapy response; machine learning; ovarian cancer; prognosis model.

Figure 1

Workflow for constructing the ERLS model.

Figure 2

64 candidate exosome-related lncRNAs. (A) A total of 2712 exosome-related lncRNAs (|Cor|≥0.4 and P<0.05). (B) 30 randomly selected lncRNAs were visualized in 840 lncRNAs. (C) 64 lncRNAs were incorporated into subsequent machine learning.

Figure 3

An ERLS was identified based on 10 machine learning algorithms and its clinical prognostic value. (A) A total of 117 algorithm combinations based on 10-fold cross-validation, the C-index of each model was calculated in the validation datasets. (B) 36 lncRNAs and their coefficients were identified based on the stepCOX (forward) combined with the Ridge algorithm. (C) Multivariate Cox regression analysis screened out 20 exosome-related lncRNAs that were independently associated with OS. (D–F). Kaplan-Meier survival analysis in the training datasets, validation datasets, and the entire TCGA cohort. (G–I) The risk score curve and the survival state heat map in the training datasets, validation datasets, and the entire TCGA cohort. (J) Univariate COX regression analysis of clinical factors and the ERLS for OS. (K) Multivariate COX regression analysis of clinical factors and the ERLS for OS.

Figure 4

The AUC and validation of the ERLS in OC patients. (A–C) The AUC of the ERLS in the training datasets, validation datasets, and the entire TCGA cohort. (D–F) The AUC of the ERLS and clinical characteristics in the training datasets, validation datasets, and the entire TCGA cohort. (G) Kaplan-Meier survival analysis in the GSE102073 dataset (log-rank test: P=0.0014). (H) The AUC of the ERLS in the GSE102073 dataset. (I) Comparison of AUC on the ERLS with other models in the entire TCGA cohort.

Figure 5

Evaluation of immune cell infiltration in high-risk and low-risk groups using xCell and ssGSEA. (A) The proportion of 64 cells in the high-risk group compared to the low-risk group was based on the xCell packages. (B) The proportion of immune cells in the high-risk group compared to the low-risk group based on the ssGSEA packages. (C) The proportion of immune cells in each OC patient. (D) The differential immune functions in the high-risk and low-risk groups. *P<0.05; **P<0.01; ***P<0.001.

Figure 6

GO and KEGG enrichment analyses of different expressed genes in high- and low-risk groups. (A) GO results. (B) KEGG results.

Figure 7

Evaluation of the immunotherapy response based on the ERLS model. (A) The TIDE score, the Exclusion score, the MSI, and the Dysfunction score. (B) Pearson’s correlation analysis between TMB and risk score. (C, D) The immunotherapy response for PD1 or CTLA4 between the high-risk group and the low-risk group. (E) The different expressions of PDL1, PD1, and CTLA4 between the high-risk group and the low-risk group. (F, G) Pearson’s correlation analysis between the expression of PDL1 or CTLA4 and the risk score. “ns” represents “not significant”. *P < 0.05, **P < 0.01.

Figure 8

The comparison of Kaplan-Meier survival analysis based on combined ERLS with PDL1 or CTLA4. (A, B) The Survival analysis for the expression of PDL1 or CTLA4. (C, D) The survival analysis was based on combined ERLS with PDL1 or CTLA4.

Figure 9

The detection of exosomes characteristics and lncRNA expression in exosomes from IOSE80 cells, SKOV3 cells, and OVCAR8 cells. (A) The results of TEM for exosomes. (B) The results of NTA for exosomes. (C) The expression of CD63 and CD81 in exosomes was detected using WB. (D–I) The expression of lncRNA in exosomes was measured using RT-PCR. The RT-PCR and WB experiments were independently repeated three times, with three replicate wells for each independent repetition. *P < 0.05, ****P < 0.0001.