© Freepik
March 23, 2026
PMC Redaktion
Detecting Prostate Cancer Early: Urologist Dr. Daniel Kaminski and Radiologist Dr. Johannes Stieß on How MRI-Guided Fusion Biopsy and AI Are Transforming Diagnosis
© PMC
With
Dr. med. Daniel Kaminski und Dr. med. Johannes Stieß
Prostate cancer is the most common cancer in men - in Germany, but also worldwide. Every year, between 65,000 and 75,000 men are newly diagnosed in this country, around 15,000 die from it. And yet the crucial question in oncology is not always whether a tumor is present, but what it means.
"Do you die from the tumor or with the tumor?" summarizes Dr. Daniel Kaminski, specialist in urology, the clinical core question during his lecture at the PMC Conference 2026. With a 5-year survival rate of over 90 percent, this is not a rhetorical exercise, but medical reality: Many men living with prostate cancer will eventually die of another cause.
The median age of onset is between the mid-60s and late 60s, but Kaminski emphasizes that the disease can start as early as between 40 and 45 years of age - which is why he consistently recommends screening from the age of 40, depending on age and family history. Anyone with a father or brother with prostate cancer should start earlier.
To understand how dangerous a tumor really is, more than a simple "benign" or "malignant" is needed. American histopathologist Donald Gleason therefore developed a system in the 1960s and 70s that is still the standard today. Kaminski explains the principle as follows: "The lower the Gleason score, the less malignant the tumor - the higher, the faster it grows."
The scale ranges from 6 to 10, with two values always being added - the most common and the second most common growth pattern in biopsy material. A Gleason score of 7 can mean 3+4 or 4+3 - and this difference is clinically relevant: 4+3 is less favorable than 3+4, even though the sum is identical.
For decades, the so-called blind biopsy was the standard procedure: With increased PSA levels and suspected tumors, tissue samples were systematically taken - without any imaging basis for where the tumor actually sits in the organ. The results were sobering. "In 40 percent of cases, significant tumors were missed, and in just as many cases, clinically insignificant tumors were detected - thus producing unnecessary findings", Kaminski describes the situation at the time.
Added to this was the phenomenon of undergrading: In 40 percent of cases, more aggressive tumor components were only discovered in the surgical specimen because the biopsy simply missed them.
Find a DoctorSearch the network for suitable specialists and contact them directly
The problem lies in the biology of the disease: Prostate carcinomas grow multifocally, that is, in several places at once. A blind needle sometimes hits more, sometimes less – and too often the wrong thing.
The way out of this dilemma came from radiology. The multiparametric MRI of the prostate – abbreviated mpMRI – has been available for about 15 years and has fundamentally changed diagnostics. Dr. Johannes Stieß, co-presenter and radiologist, explains the principle:
“The term 'multiparametric' refers to the three components that are combined in a single examination of about 20 minutes.” First, high-resolution anatomical imaging in three planes; second, diffusion imaging that measures the mobility of water molecules in the tissue; and third, perfusion analysis, which makes blood flow visible after contrast agent administration.
Particularly the second component marked the breakthrough: In tumor tissue, cell density is increased, water molecules bump into cell nuclei and membranes and can hardly move freely – this difference is registered by the MRI and made visible as a bright signal change.
“This component was the actual breakthrough for the success of prostate MRI,” says Stieß. The combination of all three parameters allows a hit rate of about 90 percent – both for the detection and exclusion of a tumor, and from a size of three millimeters.
How important this precision is in everyday life is shown by a case study from 2013: A 68-year-old patient with a PSA increase to 8.2 ng/ml and a father with prostate cancer received a conventional blind biopsy from a local urologist – result: inconspicuous.
However, the MRI had shown a highly suspicious lesion. After referral to the University Hospital Heidelberg for targeted biopsy, an aggressive tumor with Gleason score 9 and 1.3 centimeters in size was found there—exactly at the spot marked by the MRI. The blind biopsy had simply missed the area.
What clinical experience had suggested was confirmed in 2018 by the PRECISION study, published in the prestigious New England Journal of Medicine. Kaminski calls it a "true paradigm shift in urological diagnostics": Compared to blind biopsy, the MRI-guided approach detected 12 percent more clinically significant carcinomas, while 13 percent fewer clinically insignificant tumors were found—and 25 percent fewer biopsies were needed overall. Since then, the principle is: first MRI, then targeted biopsy.
This standard has now been firmly anchored in the updated S3 guideline of 2025. The recommended algorithm is: elevated PSA level – urological risk assessment – high-quality multiparametric MRI of the prostate.
To make MRI findings comparable and action-guiding, the PI-RADS classification was introduced—a five-tiered evaluation system, modeled after the BI-RADS system known from breast diagnostics. "PI-RADS indicates the likelihood of a clinically significant tumor on a scale of 1 to 5," explains Stieß. Findings in categories 1 and 2 mean no biopsy, category 3 requires an individual decision with follow-up after 9 to 12 months, while for categories 4 and 5, MRI fusion biopsy is recommended.
The centerpiece of modern prostate diagnostics is the MRI fusion biopsy—a procedure that technically merges the imaging worlds of two disciplines. The radiologist first marks the tumor and prostate contour in the MRI data set. These markings are then transferred via special software to the urologist's ultrasound device and fused with real-time ultrasound. The result: During the biopsy, the urologist can see exactly where the needle needs to go.
The procedure is performed on an outpatient basis and transrectally, with a multi-stage pain management system consisting of local anesthetic suppositories, intravenous pain medication, and a periprostatic local anesthetic. Kaminski clearly states the goal: “We want the patient to only need to be biopsied once.”
A key quality factor is the four-eyes principle. In the practice of the two specialists, it is Stieß himself who marks the tumor areas in the MRI data set – and he is personally present during the biopsy to make real-time adjustments.
“This principle significantly contributes to the high hit rate of 90 to 95 percent," says Kaminski – literature provides comparable values with 70 to 90 percent. Among the approximately 850 patients biopsied so far, significant complications were rare: six post-biopsy hemorrhages requiring hospitalization, three cases of prostatitis. At the same time, only about a quarter of all patients with elevated PSA need to be biopsied today – it used to be two-thirds.
The next step is already underway. Software from a Berlin start-up automatically analyzes MRI data, marks suspicious areas, creates the prostate contour, and prepares the data set for fusion biopsy. “Two years ago, the hit rate was around 50 percent," Stieß openly admits. Today it is already 80 percent, and the development is “unstoppable.”
In the future, AI should be able to take over the data set directly from the ultrasound machine, without manual marking. “AI is not perfect yet, but it is continuously improving – even now it significantly enhances the diagnostic workflow," says Stieß.
What remains is the realization that medical progress rarely comes from a single innovation. The revolution in prostate cancer diagnostics is the product of more precise imaging, smarter classification, interdisciplinary cooperation – and increasingly also machine learning. For the patient, this means fewer unnecessary biopsies, fewer missed diagnoses, and a significantly higher chance of finding the right tumor at the right time.
