Background and Objective: Diagnostic needle biopsies that miss clinically significant prostate cancers (PCa) likely sample benign tissue adjacent to cancer. Such samples may contain changes indicating the presence of cancer elsewhere in the organ. Our goal is to evaluate if artificial intelligence (AI) can identify morphological characteristics in benign biopsies of men with raised PSA that predict the future detection of clinically significant PCa during a 30-month follow-up. Methods: A retrospective cohort of 232 patients with raised PSA and benign needle biopsies, paired by age, year of diagnosis and PSA levels was collected. Half were diagnosed with PCa within 30 months, while the other half remained cancer-free for at least eight years. AI model performance was assessed using the area under the receiver operating characteristic curve (AUC) and attention maps were used to visualise the morphological patterns relevant for cancer diagnosis as captured by the model. Key findings and Limitations: The AI model could identify patients that were later diagnosed with PCa from their initial benign biopsies with an AUC of 0.82. Distinctive morphological patterns, such as altered stromal collagen and changes in glandular epithelial cell composition, were revealed. Conclusions and Clinical Implications: AI applied to standard haematoxylin-eosin sections identifies patients initially diagnosed as negative but later found to have clinically significant PCa. Morphological patterns offer insights into the long-ranging effects of PCa in the benign parts of the tumour-bearing organ. Patient Summary: Using AI, we identified subtle changes in normal prostate tissue suggesting the presence of tumours elsewhere in the prostate. This could aid in the early identification of potentially high-risk tumours, limiting overuse of prostate biopsies.

C.W. is on the advisory board of Navinci Diagnostics, Sweden.

This research was funded by the European Research Council via ERC Consolidator grant CoG 682810 and Technology Development project from SciLifeLab to C.W and grant no 21-1856 from the Swedish Cancer Society to A.B. The computations were enabled by the Berzelius resource provided by the Knut and Alice Wallenberg Foundation at the National Supercomputer Centre.

I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.

Yes

The details of the IRB/oversight body that provided approval or exemption for the research described are given below:

Ethical approval was granted from Regionala etikprövningsnämnden i Umeå, DNR 2010/366-31M.

I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals.

Yes

I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).

Yes

I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable.

Yes

All data produced in the present study are available upon reasonable request to the authors