EpiSwitch PSE Blood Test Reduces Unnecessary Prostate Biopsies: A Real-World Clinical Utility Study

Abstract

Background/Objectives: Prostate cancer (PCa) remains a major contributor to cancer-related morbidity and mortality worldwide. Current diagnostic strategies, largely based on PSA screening, lack specificity and sensitivity, leading to unnecessary invasive procedures and elevated healthcare costs. This real-world study evaluated the EpiSwitch^® PSE assay, a blood-based test analyzing 3D genome conformation signatures, ability to avoid unnecessary biopsies and the resulting clinical and economical benefits. Methods: 187 patients undergoing evaluation for PCa were tested with the EpiSwitch^® PSE assay. Biopsy confirmation was available for 53 patients, while predictive modeling assessed 134 patients using EpiSwitch PSE results and clinical variables. Results: Among the 187 patients evaluated, predictive modeling showed that up to 79.1% (106/134) of patients could safely defer biopsy based on a low-likelihood EpiSwitch PSE result, while an alternative model showed a 66.4% (89/134) biopsy avoidance rate. The PSE result demonstrated strong concordance with biopsy-confirmed diagnoses and was the most influential predictor in multivariate analysis, followed by PI-RADS score. The test achieved a 100% technical success rate, with an average turnaround time of 4.4 days. Conclusions: Incorporating the EpiSwitch PSE assay into clinical workflows enhances decision-making efficiency, reduces unnecessary biopsies, and improves healthcare resource utilization. These findings support the assay's strong clinical utility and economic value, highlighting its potential for broader adoption as a minimally invasive reflex test and a pre-biopsy triage tool for the early and accurate detection of prostate cancer. Future studies should include prospective, multicenter trials to confirm these results across broader populations and evaluate longitudinal outcomes of patients managed with PSE-guided care.

Keywords: 3D genome conformation; EpiSwitch PSE assay; biopsy avoidance; blood-based diagnostics; precision oncology; prostate cancer detection; real-world evidence.

Figure 1

SHAP summary plot illustrating feature importance in the predictive model for prostate cancer risk. Each dot represents a single patient, with the position on the x-axis indicating the SHAP value (impact on model output) for that feature. Features are ranked by mean absolute SHAP value, with higher values indicating greater influence on the model’s prediction. Color represents the original value of the feature (purple = low, yellow = high). In this model, the binary EpiSwitch PSE result had the strongest impact on predicted risk when combined with MRI pi_rads, showing that 79.1% of patients would be saved from biopsies.

Figure 2

SHAP summary plot illustrating feature importance in the predictive model for prostate cancer risk. This plot illustrates how each input variable influenced the model’s output prediction relative to the baseline (expected value). The base value of 1.2 represents the model’s expected prediction across the dataset, while the final output of 0.988 reflects the individualized prediction for the patient. Negative SHAP values (shown in red) decrease the predicted risk, whereas positive values (shown in yellow) increase it. In this example, a high PSA level contributed most to lowering the predicted risk (−0.141), while the PI-RADS slightly increased it (+0.086). Other features, such as 4K score, age, and PSE, had smaller marginal effects. This individualized explanation supports interpretability of the model’s output and reinforces trust in risk stratification.

Figure 3

Scaled icon array of biopsy avoidance rate of models A and B. Displayed is the predicted biopsy avoidance rate of models A and B if PSE was used as a triage tool. Each graph displays 100 icons, with each icon representing approximately 1% of the 134-patient cohort. Blue icons indicate the proportion of patients predicted to avoid biopsy, and black icons represent the rest of patients who would continue down PCa screening pathways. This figure highlights the differential impact of the two models on clinical decision-making pathways depending on how the clinician classifies Gleason 3 + 3.

Figure 4

Patient workflow under prostate cancer surveillance incorporating PSE: Illustration of a clinical decision-making pathway integrating the EpiSwitch^® PSE blood test for patients with elevated PSA or abnormal DRE findings. Based on the PSE result, patients are stratified into low or high likelihood of prostate cancer (PCa). Those with a low-likelihood result may be observed or undergo additional non-invasive testing, with reevaluation at one year. Patients with a high-likelihood result are referred for multiparametric MRI (mpMRI) and, if indicated, MRI-guided or systematic prostate biopsy. This pathway supports shared decision making and more efficient, personalized care.