PSMA PET-guided robotic biopsy beat cognitive fusion on detection and on complications

Design

Prospective, single-centre randomised clinical trial

Population

217 biopsy-naive men aged 50 to 90 with PSA 4 ng/mL or above, abnormal digital rectal examination and a PI-RADS 4 or 5 lesion

Primary outcome

Detection of prostate cancer

Effect

97.1% (101/104) PET-guided vs 81.0% (85/105) cognitive fusion, p < 0.05; complications 10.8% vs 51.4%, p < 0.01

A single-centre randomised trial enrolled 267 biopsy-naive men aged 50 to 90 with a PSA of 4 ng/mL or above and an abnormal digital rectal examination. All had multiparametric MRI, and the 217 with a PI-RADS 4 or 5 lesion were randomised to robotic-arm-assisted gallium-68 PSMA-11 PET/CT-guided biopsy (112) or MRI-directed cognitive-fusion transrectal ultrasound biopsy (105). Cancer was detected in 97.1% (101 of 104) of the PET-guided arm and 81.0% (85 of 105) of the cognitive-fusion arm, p < 0.05. For PI-RADS 5 lesions detection was 100% against 95.1% (p = 0.09). Complications occurred in 10.8% against 51.4% (p < 0.01), with all five major complications in the cognitive-fusion arm; median pain score was 3 against 5, procedures were shorter, and core positivity was higher (60% ± 20%) despite fewer cores.

The comparator is what to hold in mind. Cognitive fusion means the operator carries the MRI finding in their head and aims for it under ultrasound — it is the commonest approach where software fusion is unavailable, and it is the weakest of the targeted techniques. This trial does not compare PSMA PET guidance with software-fusion or in-bore MRI biopsy, so it establishes superiority over the cheapest option rather than over the best one.

Even so, the complication difference is striking and hard to explain away: 51.4% against 10.8% is not a marginal result, and it points as much at the transrectal route and the number of cores as at the targeting method. This is also an Indian trial, from a centre with both a PET service and a robotic biopsy arm, which makes it directly relevant here — and simultaneously flags the limitation, because that combination exists in very few Indian centres. Where PSMA PET is already being done for staging, integrating the biopsy into the same visit is an argument worth examining; where it is not, this is not a reason to add a PET scan to a diagnostic pathway.

• Note the comparator was cognitive fusion, not software fusion or in-bore MRI biopsy.
• The complication difference, 51.4% against 10.8%, likely reflects route and core number as well as targeting.
• Consider integrating biopsy with PET where a PSMA study is already planned for staging.
• Do not add a PSMA PET to a diagnostic pathway on the strength of a single-centre trial.
• Detection in PI-RADS 5 lesions was already 95.1% with cognitive fusion — the gain is concentrated in PI-RADS 4.

The statistics, in plain English

The headline gap, 97.1% against 81.0%, is a detection rate and not an accuracy: with no independent reference standard applied to every man, it cannot separate finding more cancer from finding more of the indolent kind. The PI-RADS 5 subgroup is the informative one — 100% against 95.1%, p = 0.09 — because when the comparator is already near ceiling there is nothing to gain, which locates the whole benefit in PI-RADS 4 lesions where targeting is harder. The complication comparison is the most robust result here, since it is an objective count and the difference is large. Note that arm 1 patients with negative PET went on to cognitive-fusion biopsy, so the two arms are not cleanly independent for every outcome.

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.