To address this, we conducted a multicenter ambispective biomarker study including 350 patients with mHSPC. PTEN mRNA expression was measured using nCounter in archival FFPE tumor samples, and RNA-Seq was used to characterize transcriptional programs associated with PTEN status. PTENlow was defined by a validated cut-off (Jiménez, Eur Urol Onc 2024) and correlated with survival outcomes. We also derived and externally validated a PTEN-low related signature using microarray data from the CHAARTED trial.
We found that PTENlow tumors, observed in 28.2% of patients, were independently associated with shorter castration-resistant prostate cancer–free survival (CRPC-FS) and overall survival (OS). This was also validated in patients from the CHAARTED trial, where low PTEN expression correlated with shorter CRPC-FS and OS, and did not seem to benefit from the addition of docetaxel to androgen deprivation therapy in terms of OS.
At the molecular level, PTENlow tumors showed activation of cell cycle, DNA repair, chromosomal instability, and neuroendocrine programs, along with significant suppression of androgen receptor signaling. Immune pathway analysis revealed decreased expression of T-cell activation and effector pathways, with high myeloid suppressor cells, consistent with an immunosuppressive or “immune-excluded” phenotype. Among the transcriptional targets, EZH2 emerged as significantly upregulated, aligning with its known role in lineage plasticity and neuroendocrine dedifferentiation, and suggesting a rationale for therapeutic targeting in PTEN-altered disease. Patients with high EZH2 expression together with low PTEN expression exhibited the shortest OS and CRPC-FS-
To better capture the biological “PTEN-low” phenotype, we developed a transcriptomic signature using data from the CHAARTED clinical trial. An elastic-net model identified a 39-gene signature distinguishing PTENlow tumors. This signature showed good performance (AUC 0.75) and was strongly correlated with PI3K/AKT/mTOR, IL2, and interferon-alpha pathways, consistent with known biology. In the CHAARTED cohort, higher PTEN-low signature expression was independently associated with shorter CRPC-free survival and overall survival. We then validated the 39-gene set in our RNA-Seq dataset (n = 60). While its predictive accuracy for PTENlow status was modest (AUC 0.61) and no significant survival associations were observed, the signature again correlated with PI3K/AKT/mTOR, interferon-alpha, and IL2-STAT5 pathways, reproducing the biological signals from the training cohort. The relevance of this multi-gene signature is that it captures the true PTEN-loss phenotype, potentially offering greater biological and clinical relevance than relying on PTEN mRNA threshold alone.
Our study does face limitations, including its retrospective and non-randomized design, possible selection bias, missing data, and treatment heterogeneity across first-line therapy groups like triplets. In addition, the prognostic value of the PTEN-low-related signature remains unconfirmed, and larger RNA-Seq cohorts will be needed to establish its clinical relevance.
In conclusion, low PTEN expression is a strong adverse prognostic factor in mHSPC, linked to a unique molecular profile and an immune tumor microenvironment. Our findings matter because they show that low PTEN expression defines a subset of mHSPC, with potential targetable pathways, highlights opportunities for rational therapeutic combinations, including AKT inhibitors, epigenetic modulators, and novel immune-based strategies. Should we treat these patients with triplet therapy? Are AKT inhibitors enough, or should we combine them with other treatment strategies? We believe that integrating PTEN status and PTEN-loss signatures into clinical and translational research may help the development of personalized treatment strategies for patients with PTEN-altered prostate cancer.
Written by: Marta Garcia de Herreros,1-4 Natalia Jiménez,1 Begoña Mellado1-4
- Translational Genomics and Targeted Therapeutics in Solid Tumors Lab, Fundació de Recerca Clínic Barcelona-Institut d'Investigacions Biomèdiques August Pi i Sunyer (FRCB-IDIBAPS), Barcelona, Spain.
- Medical Oncology Department, Hospital Clínic, Barcelona, Spain.
- Uro-Oncology Unit, Hospital Clínic, University of Barcelona, Barcelona, Spain.
- Department of Medicine, University of Barcelona, Barcelona, Spain.