Interplay between tumor mutation burden and the tumor microenvironment predicts the prognosis of pan-cancer anti-PD-1/PD-L1 therapy
Bibliographic record
Abstract
Introduction: Immune checkpoint inhibitors (ICIs) have revolutionized the treatment landscape for advanced cancers, yet their efficacy remains heterogeneous among patients. Tumor mutation burden (TMB) has been extensively explored as a potential biomarker for predicting ICI response. However, its application is limited by several factors, including inconsistent predictive power across different tumor types and the lack of a clear relationship with overall survival (OS). This study aimed to explore the complex interplay between TMB and the tumor microenvironment (TME) and to identify novel predictive biomarkers that can enhance the precision of ICI therapy across multiple cancer types. Methods: We systematically collected and analyzed genomic and clinical data from patients receiving anti-PD-1/PD-L1 immunotherapy across multiple cohorts. Our dataset included information from The Cancer Genome Atlas (TCGA) pan-cancer database and various ICI clinical trials. We first screened immunosuppression-related genes (ISRGs) that might interfere with TMB's predictive role by analyzing the survival data and gene expression profiles of patients. Using LASSO regression and multivariable Cox proportional hazards analysis, we constructed a risk model based on these ISRGs. The model's predictive ability was rigorously validated in multiple independent cohorts. Additionally, we employed algorithms such as CIBERSORT and ESTIMATE to assess the correlation between the risk score and TME components. To further explore the therapeutic implications of our findings, we focused on RPLP0, a ribosomal protein that emerged as a robust biomarker in our model. We investigated its expression in tumor tissues and evaluated the impact of its knockdown on immunotherapeutic efficacy using in vitro and in vivo experiments. Results: Our comprehensive analysis revealed that the predictive power of TMB varies significantly across different cancer types and is highly dependent on its interaction with the TME. In tumors with a favorable immune microenvironment, characterized by high CD8+ T cell infiltration and M1 macrophage presence, TMB maintained its predictive ability. However, in immunosuppressive microenvironments, TMB alone failed to accurately predict patient outcomes. We identified 304 ISRGs and developed a 10-gene risk signature that demonstrated reliable prognostic predictive ability in both ICI cohorts and TCGA pan-cancer. The risk score derived from this model was significantly associated with stromal components and an immunosuppressive TME, characterized by elevated levels of M0 macrophages and activated mast cells. Notably, RPLP0 was identified as the most robust predictive marker during model building. We demonstrated its abnormal overexpression in tumor tissues and further showed that intratumoral RPLP0 knockdown in a subcutaneous bladder cancer model could enhance the efficacy of immunotherapy. The combination of RPLP0 knockdown and anti-PD-1 treatment resulted in significantly suppressed tumor growth and prolonged survival in mice, accompanied by elevated levels of IFN-γ and TNF-α in serum samples, indicating enhanced anti-tumor immunity. Conclusion: This study establishes a reliable risk model that complements TMB in guiding treatment decisions for ICI therapy. By incorporating the interaction between TMB and the TME, our model provides a more accurate prediction of patient prognosis and treatment response across multiple cancer types. The risk score's association with immunosuppressive TME components underscores the importance of considering the tumor's microenvironment in treatment planning. Furthermore, our findings highlight RPLP0 as a promising therapeutic target for combination immunotherapy. The robust predictive ability of our model across various cohorts and its potential to improve therapeutic outcomes offer new insights and directions for enhancing the efficacy of ICI therapy. Future research should focus on further validating this model in larger and more diverse cohorts, refining the gene set selection process, and exploring the specific mechanisms through which the identified biomarkers influence the TME and treatment response.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.003 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".