Design: Systematic review and diagnostic meta-analysis
Population: 16 studies, 1,999 men with intermediate- or high-risk prostate cancer
Primary outcome: Sensitivity and specificity for histologically confirmed pelvic nodal metastases
Effect: PSMA PET/CT sensitivity 54.6%, specificity 92.7%; MRI sensitivity 33.0%, specificity 94.9%

A systematic review of 16 studies with 1,999 men with intermediate- or high-risk prostate cancer compared PSMA PET/CT and multiparametric MRI against histology from pelvic lymph node dissection.

Pooled sensitivity for nodal metastases was 54.6% for PSMA PET/CT and 33.0% for MRI, with specificity above 92% for both. PSMA PET/CT appeared more sensitive, but the comparison is indirect, drawn from largely separate groups of studies, and the MRI estimate rests on only six studies with a wide confidence interval.

The practical point is in the reports. False-positive nodes were uncommon, but a negative PSMA PET/CT does not exclude microscopic nodal disease, and the reports radiologists write should not imply that it does. It was published in August 2026.

Report a negative PSMA PET/CT as no PSMA-avid nodes, not as no nodal disease.
Expect PSMA PET/CT to miss small-volume nodal metastases; sensitivity was about 55%.
Specificity was over 90%, so false-positive nodes were uncommon in these surgical series.
Discuss the limits of imaging when urologists are deciding on lymph node dissection.

Why it matters

Overconfidence in a negative scan can lead to under-staging.

The statistics, in plain English

A pooled sensitivity of 54.6% (95% CI 43.7–65.1%) means PSMA PET/CT detected about half of pathologically confirmed pelvic nodal metastases. The MRI interval is very wide (18.7–51.3%) because few studies contributed. An indirect comparison between separate study groups is weaker than a head-to-head study.