Stockholm3 with MRI-Targeted Biopsy Noninferior to PSA for Prostate Cancer Screening

Combining Stockholm3 blood test and MRI-targeted biopsies is noninferior to prostate-specific antigen for detecting clinically significant prostate cancer and reduced over-detection, a new study found.

The population-based, randomized, open-label noninferiority trial (STHLM3-MRI), published in The Lancet Oncology, included 2,293 men ages 50 to 74 with elevated risk of prostate cancer based on a prostate-specific antigen (PSA) of 3 ng/mL or higher or a Stockholm3 score of 0.11 or higher.

Among the participants, 1,372 were assigned to the experimental group, receiving biparametric MRI followed by MRI-targeted and systematic biopsy in MRI-positive participants, and 921 were assigned to the standard group, receiving systematic prostate biopsies.

“With this research, we provide level 1 evidence on the use of blood-based risk prediction (i.e. the Stockholm3 test) and MRI for prostate cancer screening,” principal investigator Tobias Nordström, MD, PhD, a consultant urologist and associate professor at Karolinska Institutet in Sweden, told Diagnostic Imaging. “We show a way of maintaining detection of clinically significant disease, while simultaneously decreasing the pivotal problem of over-detection drastically.”

The primary outcome was the detection of clinically significant prostate cancer at prostate biopsy, defined as a Gleason score of 3 + 4 or higher. Secondary outcomes included the proportion of men with clinically insignificant prostate cancer, defined as a Gleason score of 3+3. Comparisons were made using Stockholm3 scores of 0.11 and 0.15 as cutoffs

A Stockholm3 of 0.15 or higher showed identical detection of clinically significant prostate cancer to a PSA of 3 ng/mL or higher and led to 36% fewer MRI procedures (545 vs 846; 0.64 [0.55–0.82]) and 8% fewer biopsies (311 vs 338; 0.92 (0.86–1.03).

A Stockholm3 of 0.11 or higher was noninferior to PSA for detection of clinically significant prostate cancer (227 vs 192; relative proportion [RP] 1·18 [95% CI 1·09–1·28], p<0·0001 for non-inferiority), detected a similar number of clinically insignificant prostate cancers (50 vs 41; 1·22 [0·96–1·55], p=0·053 for superiority) and was associated with more MRIs and biopsies.

When combined with MRI-targeted and systematic biopsies, Stockholm3 of 0.11 or higher was associated with higher detection of clinically significant cancers than screening using PSA and systematic biopsies (227 [3·0%] men tested vs 106 [2·1%] men tested; RP 1·44 [95% CI 1·15–1·81]), lower detection of low-grade cancers (50 [0·7%] vs 73 [1·4%]; 0·46 [0·32–0·66]), and fewer biopsies.

Patients in the experimental group also had a lower incidence of prescription of antibiotics for infection and a lower incidence of admission to hospital.

“High-quality knowledge takes endurance and resources- these data are the result of dedicated work by our research team during 10 years,” Nordström said. “We are so grateful also to all funders and participants.

“Our results also illustrate the profound difference between clinical and screening cohorts. For example, the distribution of disease risk as expressed by median PSA is lower in our trial (using a screening-population) than in the previous Precision trial (median PSA 4.3 versus 6.8). Therefore, any diagnostic tools suggested for use in screening contexts needs to be evaluated in such cohorts.”

Research is ongoing, including the final states of health economic evaluations and a second screening round for participants in STHLM3-MRI.

“In upcoming analyses, we will present how use of the Stockholm3 test and MRI can be optimized for repeat screening,” Nordström said.