Identification of disulfidptosis-related molecular subtypes in pleural mesothelioma demonstrates a correlation with the immune infiltration, prognosis and therapy efficacy

Our plain-English summary of Identification of disulfidptosis-related molecular subtypes in pleural mesothelioma demonstrates a correlation with the immune infiltration, prognosis and therapy efficacy, by J. Ding, X. Huang, W. Yao, et al (Discover Oncology, 2026).

The study

• Authors: J. Ding, X. Huang, W. Yao, T. Lin
• Journal: Discover Oncology
• Published: 2026;17(1)
• Read it: PubMed · Publisher · Free full text

Abstract

Background: Pleural mesothelioma (PM) is a rare and aggressive malignancy with increasing incidence, high mortality, and limited treatment options. Disulfidptosis, a newly identified form of regulated cell death, may play an important role in cancer, but its relevance in PM remains unclear.

Methods: Transcriptomic and clinical data of PM were obtained from public databases. PM was classified into molecular subtypes based on disulfidptosis-related genes (DRGs). We then compared prognosis, tumor microenvironment, metabolic features, predicted response to immune checkpoint blockade, and chemotherapy sensitivity between subtypes. In addition, a prognostic signature was established, and key genes were identified using random survival forest analysis.

Results: Two novel DRG-defined molecular subtypes were identified in PM. The DRGs-high subtype (Cluster 2) was associated with significantly worse overall survival, an immunosuppressive microenvironment dominated by M2 macrophages, and a lower predicted likelihood of response to immune checkpoint blockade. In contrast, the DRGs-low subtype (Cluster 1) showed metabolic reprogramming toward glycolysis and fatty acid oxidation, increased NK cell infiltration, and reduced sensitivity to chemotherapy. An 11-gene prognostic signature demonstrated predictive performance, with an AUC of 0.95 for 5-year survival (95%CI 0.881-1.013). Random survival forest analysis identified LMNB2 and CDCA2 as key genes, both of which were highly expressed in tumor tissues.

Conclusions: PM can be classified into two DRG-defined molecular subtypes with distinct prognostic, immune, and metabolic features. These findings provide insights into PM heterogeneity and may support future studies on therapeutic stratification. The prognostic signature remains exploratory and requires validation in independent external cohorts before clinical application.

© 2026. The Author(s).

In plain English

This was a computer analysis of public gene-activity and clinical data. It did not study asbestos exposure. The authors sorted 87 pleural mesothelioma tumor samples from The Cancer Genome Atlas into two groups by the activity of 15 genes linked to disulfidptosis, a recently described form of cell death. Cluster 1 had 41 samples. Cluster 2 had 46 samples and higher activity of these genes. Patients in Cluster 1 had better overall, disease-specific and progression-free survival. Computer estimates suggested that Cluster 2 tumors contained more M2 macrophages and might respond less well to immune checkpoint drugs. Cluster 1 was estimated to be less sensitive to all nine chemotherapy and targeted drugs assessed. Cluster 2 also had more non-epithelioid tumors (21 of 46, versus 9 of 41). An 11-gene risk score gave a hazard ratio of 4.820 (95% CI 2.812 to 8.260) for overall survival in the high-risk group compared with the low-risk group (43 patients each). The authors concluded that the score may support future risk stratification but needs independent external validation. They noted the small cohort and possible overfitting.

We wrote this plain-English summary from the full article with the help of AI, and a second AI model checked each sentence against the article. It is not the authors' wording; the abstract and the article are their own account.

What the study did and found

• Design: Bioinformatics (computer) analysis of publicly available tumor gene-expression and clinical data; no new patients or experiments
• Who was studied: 87 pleural mesothelioma tumor samples from the public TCGA-MESO dataset. Table 1 lists 71 male and 16 female patients. Earlier comparisons also covered 33 cancer types in TCGA. Location and dates of sample collection are not given.
• Exposure: Not an exposure study; asbestos exposure was not assessed. Tumors were grouped by expression of 15 disulfidptosis-related genes.
• Main results:
  • The analysis identified two subtypes: Cluster 1 (41 samples) and Cluster 2 (46 samples). The disulfidptosis-related genes were more highly expressed in Cluster 2.
  • Patients in Cluster 1 had better overall survival, disease-specific survival and progression-free survival than patients in Cluster 2.
  • Table 1: non-epithelioid mesothelioma made up 9 of 41 samples in Cluster 1 and 21 of 46 in Cluster 2 (P 0.020).
  • Computer estimates showed more memory B cells and activated NK cells in Cluster 1, and more M0 and M2 macrophages in Cluster 2. Table 1 gives M2 macrophages as 0.32 plus or minus 0.10 in Cluster 1 versus 0.38 plus or minus 0.11 in Cluster 2 (P 0.009).
  • The TIDE prediction tool suggested that patients in Cluster 2 may respond less well to immune checkpoint blockade than those in Cluster 1 (contingency table p = 0.004). This was a prediction, not observed treatment response.
  • Drug sensitivity was estimated from GDSC database data, not tested in patients. Cluster 1 had higher estimated IC50 values, meaning lower predicted sensitivity, for all nine drugs: paclitaxel, docetaxel, vinorelbine, cisplatin, gemcitabine, belinostat, vorinostat, crizotinib and pazopanib.
  • An 11-gene risk score split patients into high-risk (n = 43) and low-risk (n = 43) groups. The high-risk group had poorer overall survival: HR = 4.820, 95% CI 2.812 to 8.260, p less than 0.0001.
  • In the same TCGA data used to build the risk score, the 5-year AUC was 0.95 (95% CI as printed 0.881 to 1.013).
  • In univariate and multivariate Cox analyses that included age, sex and stage, the risk score was an independent prognostic factor. A nomogram combining clinical data and the risk score had AUC values of 0.771, 0.793 and 0.872 at 1, 2 and 3 years.
  • A random survival forest ranked LMNB2 and then CDCA2 as the most important genes in the model. Their higher expression in cancer versus normal tissue was shown in other cancer types (lung, breast, esophageal, colon, cervical, liver), not in mesothelioma tissue.
• The authors' conclusion: The authors concluded that the prognostic signature may support future risk stratification in pleural mesothelioma but requires independent external validation before clinical use. They described the study as a hypothesis-generating framework.
• Limitations the authors state:
  • The prognostic model was built from a relatively small TCGA-MESO cohort and had no independent external validation.
  • Overfitting of the model could not be fully excluded, so the AUC estimates should be interpreted cautiously.
  • The subtypes rest on gene-expression patterns, not direct evidence that disulfidptosis occurs. Disulfide accumulation, NADPH depletion and actin collapse were not measured.
  • Bulk transcriptomic data could not show how cells are distributed or organized within the tumors.
  • The links between LMNB2 and CDCA2 and prognosis are correlational.
  • The treatment-response framework linking subtypes to specific immunotherapy approaches is hypothetical and needs prospective validation.

Topics

Medical terms in this study, explained in our Lexicon: Pleural Mesothelioma, Prognosis, Tumor Microenvironment, Chemotherapy, Overall Survival, Macrophage.

Related studies

• Analysis of the correlation between Zeste enhancer homolog 2 (EZH2) mRNA expression and the prognosis of mesothelioma patients and immune infiltration — K. Fan, C.L. Zhang, et al. (2022)
• A Mitophagy-Related Gene Signature for Subtype Identification and Prognosis Prediction of Hepatocellular Carcinoma — C. Liu, Z. Wu, et al. (2022)
• Systemic characterization of alternative splicing related to prognosis and immune infiltration in malignant mesothelioma — J. Lai, H. Yang, et al. (2021)
• Identification of Tumor Antigens and Immune Subtypes of Malignant Mesothelioma for mRNA Vaccine Development — S. Wang, Y. Yang, et al. (2022)
• Novel combinatorial therapy of oncolytic adenovirus AdV5/3-D24-ICOSL-CD40L with anti PD-1 exhibits enhanced anti-cancer efficacy through promotion of intratumoral T-cell infiltration and modulation of tumour microenvironment in mesothelioma mouse model — M. Garofalo, M. Wieczorek, et al. (2023)
• Unraveling tumor microenvironment heterogeneity in malignant pleural mesothelioma identifies biologically distinct immune subtypes enabling prognosis determination — K. Yang, T. Yang, et al. (2022)

Browse more research in Discover Oncology, or return to medical research.