Prostate Clarity: A New Noninvasive Biomarker for Prostate Cancer Screening
AI-identified biomarker-based test reduces need for prostate biopsies in men with high PSA levels

Prostate cancer remains a significant health challenge, with prostate-specific antigen (PSA) testing failing to accurately distinguish between malignant prostate cancers and benign conditions with similar symptoms, such as benign prostatic hyperplasia (BPH), resulting in 70 percent of prostate biopsies being negative. Recent scientific advancements highlight the potential of using artificial intelligence (AI) to identify novel biomarkers for diagnostic test development (as well as therapeutics design). The development of such biomarkers into diagnostic tests could enhance diagnostic precision and reduce unnecessary medical procedures.
Prostate cancer rates are rising
Prostate cancer is one of the most common cancers among men worldwide, with approximately 1.4 million new cases and nearly 400,000 deaths annually. The incidence of prostate cancer has seen a significant rise in recent years, partly due to increased PSA screening and early detection efforts. Although testing for prostate cancer early and often has improved cancer survival rates, the challenge lies in accurately distinguishing between BPH and aggressive prostate cancer.
BPH is very common, affecting around 50–60 percent of men in their 60s and up to 80–90 percent of those over 70, and presents with many of the same symptoms, such as difficulty urinating and an enlarged prostate.
Current noninvasive testing methods for prostate cancer fall short
PSA testing has been the cornerstone of prostate cancer screening for decades. While it has contributed significantly to early detection, the PSA test is not without its limitations. Elevated PSA levels can be indicative of both prostate cancer and BPH, leading to a high rate of false positives. This ambiguity often results in unnecessary biopsies, which are invasive, costly, painful, and can cause complications such as infections and bleeding. Moreover, the psychological impact of undergoing repeated biopsies can be substantial for patients and can serve as a deterrent for additional screenings.
Filamin A: a novel protein-based biomarker for cancer detection
Filamin A is a cytoskeletal protein involved in various cellular processes, including maintaining cell shape, motility, and signal transduction. Its usefulness as a biomarker in prostate cancer detection was identified by BPGbio’s proprietary NAi platform using robust systems medicine models in AI. These models were based on longitudinal, biological data and formulated using a causal, hypothesis-free approach. This allowed for AI to have a more complete understanding of human biology and to organically point to biomarkers of interest.
Our studies and subsequent clinical trials of our pstateDx test confirmed that filamin A levels in serum could effectively differentiate between benign prostatic hyperplasia and prostate cancer, providing a more specific indicator than traditional PSA tests. The test was shown to be effective across various ethnicities.
Filamin A is the first-ever AI-discovered protein-based oncology biomarker, and it underscores the potential of AI-powered platforms in the discovery and development of diagnostic tools. And while there are certainly more biomarkers ready to be identified using AI, researchers must be aware of their approach, ensuring that models are built with robust data that considers a more multifaceted understanding of the genomics, proteins, lipids, and more.