PET imaging of new target CDK19 in prostate cancer.

Purpose: Prostate-specific membrane antigen (PSMA)-positron emission tomography (PET) is a superior method to predict patients' risk of cancer progression and response to specific therapies. However, its performance is limited for neuroendocrine prostate cancer (NEPC) and PSMA-low prostate cancer cells, resulting in diagnostic blind spots. Hence, identifying novel specific targets is our aim for diagnosing those prostate cancers with low PSMA expression.

Methods: The Cancer Genome Atlas (TCGA) database and our cohorts from men with biopsy-proven high-risk metastatic prostate cancer were used to identify CDK19 and PSMA expression. PDX lines neP-09 and P-16 primary cells were used for cellular uptake and imaging mass cytometry in vitro. To evaluate in vivo CDK19-specific uptake of gallium(Ga)-68-IRM-015-DOTA, xenograft mice models and blocking assays were used. PET/CT imaging data were obtained to estimate the absorbed dose in organs.

Results: Our study group had reported the overexpression of a novel tissue-specific gene CDK19 in high-risk metastatic prostate cancer and CDK19 expression correlated with metastatic status and tumor staging, independently with PSMA and PSA levels. Following up on this new candidate for use in diagnostics, small molecules targeting CDK19 labeled with Ga-68 (Ga-IRM-015-DOTA) were used for PET in this study. We found that the Ga-IRM-015-DOTA was specificity for prostate cancer cells, but the other cancer cells also took up little Ga-IRM-015-DOTA. Importantly, mouse imaging data showed that the NEPC and CRPC xenografts exhibited similar signal strength with Ga-IRM-015-DOTA, but Ga-PSMA-11 only stained the CRPC xenografts. Furthermore, target specificity was elucidated by a blocking experiment on a CDK19-bearing tumor xenograft. These data concluded that Ga-CDK19 PET/CT was an effective technology to detect lesions with or without PSMA in vitro, in vivo, and in the PDX model.

Conclusion: Thus, we have generated a novel PET small molecule with predictive value for prostate cancer. The findings indicate that Ga-CDK19 may merit further evaluation as a predictive biomarker for PET scans in prospective cohorts and may facilitate the identification of molecular types of prostate cancer independent of PSMA.