This study aimed to systematically assess how androgen deprivation therapy (ADT) affects the immune landscape of prostate cancer at multiple biological levels. The investigators combined clinical immune monitoring, analysis of patient-derived xenograft (PDX) transcriptomes, and single-cell RNA sequencing (scRNA-seq) of tumor tissues to characterize ADT-mediated remodeling of the tumor immune microenvironment (TIME).
Three complementary approaches were used.
Clinical cohort: Twelve prostate cancer patients treated with ADT at Peking University People’s Hospital underwent dynamic monitoring of serum prostate-specific antigen (PSA), serum testosterone, peripheral blood immune-cell proportions, and cytokine expression changes.
PDX transcriptomic analysis: Publicly available transcriptome data (GEO dataset GSE41193) from five patient-derived tumor xenograft models were analyzed to compare gene expression before and after castration. Enrichment and immune infiltration–related bioinformatics analyses were performed on differentially expressed genes.
Single-cell RNA sequencing: Four prostate cancer tissue samples were profiled by scRNA-seq — two samples from patients who received neoadjuvant ADT and two untreated samples — to examine intratumoral cell composition and T-cell functional states.
In the clinical cohort, ADT produced the expected biochemical responses: serum PSA and testosterone levels decreased significantly. However, ADT did not produce marked changes in peripheral blood immune-cell proportions or most circulating cytokines according to the reported measurements. The abstract indicates no significant systemic immune perturbation in peripheral blood despite clear endocrine effects.
Analysis of the five PDX transcriptomes showed that genes differentially expressed after castration were enriched in neural-related and immune-regulatory pathways. The authors report that characteristics of immune-cell infiltration exhibited trend-level changes across multiple immune cell types in the PDX models, suggesting that castration-induced transcriptional shifts include immune-relevant signals. The abstract does not provide detailed lists of specific immune-cell populations altered nor effect sizes from the PDX analysis.
scRNA-seq of four tumor specimens revealed intratumoral changes associated with ADT. Tumors from patients who had received neoadjuvant ADT showed a decreased proportion of intratumoral effector CD8+ T cells compared with untreated tumors. At the gene-expression level, cytotoxicity-associated genes — including GZMA, GZMB, GNLY, and NKG7 — were downregulated in the post-ADT tumor T-cell compartment. Conversely, regulatory T cells (Tregs) were relatively enriched in the ADT-treated tumors, indicating a shift toward immunosuppressive cellular composition within the tumor.
Hallmark pathway analysis performed on the single-cell data indicated suppression of interferon signaling pathways after ADT. Taken together with the downregulation of cytotoxic effector genes and the relative increase in Tregs, these findings point to functional impairment of intratumoral T-cell cytotoxicity and an enrichment of immunosuppressive programs in the TIME following ADT.
Across datasets and methods, the study identifies a consistent pattern: while ADT produces expected systemic biochemical changes (PSA and testosterone reduction), it is associated at the tumor level with features suggestive of immune suppression — decreased effector CD8+ T-cell representation and cytotoxic gene expression, enrichment of Tregs, and downregulated interferon signaling. The authors interpret these findings to mean ADT may promote an immunosuppressive TIME in prostate cancer.
Clinically, this pattern provides a theoretical rationale for combining ADT with immune-modulating therapies to counteract or reverse ADT-associated immunosuppression. The results suggest that timing, sequencing, or addition of agents that enhance T-cell cytotoxicity or reduce regulatory/suppressive cell functions could be important when designing combination regimens.
The abstract presents descriptive and integrative findings but omits many details necessary to fully evaluate strength and generalizability: it does not report detailed statistical values, effect sizes, longitudinal timing of sampling relative to ADT, patient clinical stages, or duration of ADT exposure. The PDX analysis is based on five cases (GSE41193) and the scRNA-seq cohort comprises four tumors (two ADT-treated, two untreated), which are small sample sizes. The abstract does not report whether peripheral blood immune readouts were correlated with intratumoral changes or clinical outcomes. These details were not reported in the abstract.
The study's multi-level data indicate that ADT in prostate cancer reduces PSA and testosterone systemically but is associated with a trend toward an immunosuppressive tumor microenvironment characterized by weakened T-cell killing function and relative enrichment of immunosuppressive cells such as Tregs, alongside suppression of interferon pathways. These preliminary findings support further investigation into ADT plus immune-modulating strategies and warrant larger, detailed studies to define mechanisms, timing, and therapeutic approaches.