PSA density should set the follow-up interval after a negative prostate MRI

Design

pre-specified sub-analysis of a population-based randomised screening trial, Kaplan-Meier and log-rank analysis

Population

1,685 men with a negative initial MRI (PI-RADS ≤2) in the Göteborg-2 trial, median follow-up 5.5 years

Primary outcome

cumulative incidence of clinically significant prostate cancer (Gleason ≥3+4)

Effect

7-year incidence 5.5% at PSA density <0.10, 13.8% at 0.10–0.15 and 11.9% at ≥0.15 (P < 0.001 for lowest vs both); modelled strategy cuts MRI 60% and biopsies 46%, delaying diagnosis in 17%

This pre-specified sub-analysis of the population-based randomised Göteborg-2 screening trial took 1,685 men who had PSA testing between October 2015 and June 2021 and a negative initial MRI (PI-RADS 2 or below), and followed them to September 2024 — a median of 5.5 years. Men were grouped by baseline PSA density below 0.10, 0.10 to under 0.15, and 0.15 ng/mL² or above.

One hundred and sixty cancers were diagnosed, 80 clinically significant (Gleason 3+4 or above), including 22 at Gleason 4+3 or above. Seven-year cumulative incidence of significant cancer was 7.1 per cent overall: 5.5 per cent below density 0.10, 13.8 per cent in the middle band and 11.9 per cent at 0.15 or above (P < 0.001 for the lowest band against both higher ones). Eighty-seven per cent of repeat MRIs remained negative.

The modelled strategy is where this becomes actionable. Biennial PSA with repeat MRI triggered only when density reaches 0.10 would cut MRI use by 60 per cent and biopsies by 46 per cent, and halve detection of clinically insignificant cancer. The cost is explicit and should be quoted with the benefit: significant cancer diagnosis delayed in 17 per cent of cases, including 10 men with Gleason 3+4 and one with Gleason 4+3 or above. For a screening programme that trade is defensible. The other number to carry is that 42 significant cancers, including 12 at Gleason 4+3 or above, occurred in the low-density group — so low density buys a longer interval, not discharge.

• Set the post-negative-MRI follow-up interval by PSA density rather than by PSA alone
• Use density 0.10 ng/mL² as the trigger for repeating MRI, with biennial PSA otherwise
• Do not discharge men with low density — significant cancers still arose in that group
• Quote the trade-off honestly when explaining a longer interval: fewer scans and biopsies, some diagnoses later
• Note that 87 per cent of repeat MRIs were negative, which is the waste this strategy targets

Why it matters

It converts a negative MRI from a single reassurance into a risk-stratified follow-up interval.

Don't overread it

The 60 per cent and 46 per cent reductions are modelled on this cohort, not the results of a trial comparing the two follow-up strategies.

The statistics, in plain English

The middle band (13.8 per cent) came out higher than the top band (11.9 per cent), which is what small numbers look like — the reliable contrast is the lowest band against both of the others, and that is the one the authors tested. The efficiency figures are modelled from this cohort rather than observed in a trial of the strategy, so they describe what would have happened here, not what will happen elsewhere.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.