medigraphic.com
Organo Oficial de la Sociedad Mexicana de Urología

2026, Number 3

Next >>

Rev Mex Urol 2026; 86 (3)

Artificial intelligence versus multidisciplinary teams in high-risk prostate cancer with positive surgical margins: a pilot proof-of-concept study

Duarte-Valdés R, Lagos-Beitz HE, Martínez-Delgado GH,Castillejos-Molina RA

Language: English
References: 18
Page: 1-15
PDF size: 318.47 Kb.


ABSTRACT

Objective: to compare treatment recommendations from multidisciplinary teams (MDTs) and artificial intelligence (AI) in high- and very high-risk prostate cancer (PCa) with positive surgical margins (PSM) after radical prostatectomy (RP).
Design, methodology or approach: we conducted a retrospective proof-ofconcept study at a tertiary referral center. Among 700 men who underwent RP between 2010–2024, 76 had PSM, and 14 had documented preoperative MDT discussions. Five patients fulfilled inclusion criteria: high-/very high-risk PCa (per NCCN), confirmed PSM, and both pre- and postoperative MDT evaluations. Clinical and pathological data were entered into ChatGPT-4.0 (OpenAI), primed with NCCN, AUA, and EAU guidelines. The AI was queried in a zero- shot manner for treatment recommendations, and concordance was evaluated using Cohen’s kappa (κ).
Results: five patients were included in the final analysis. AI recommendations aligned with MDT decisions in 2 of 5 cases (40 %) both preoperatively and postoperatively. Agreement was minimal for initial therapy (κ = 0.00) and slight for postoperative management (κ = 0.12).
Limitations or implications: this pilot study is limited by its small sample size and retrospective design, restricting generalizability.
Originality or value: this is the first study to directly compare AI-driven and MDT recommendations in high-risk PCa with PSM in Mexico.
Findings or conclusions: AI generated rapid, guideline-based recommendations, but showed limited concordance with MDT decisions in this pilot cohort. AI may support efficiency and guideline adherence as an adjunctive tool; however, this study did not evaluate whether either recommendation strategy leads to superior clinical outcomes.


REFERENCES

  1. Bray F, Laversanne M, Sung H, Ferlay J,Siegel RL, Soerjomataram I, et al. Globalcancer statistics 2022: GLOBOCAN estimatesof incidence and mortality worldwide for 36cancers in 185 countries. CA: a cancer journalfor clinicians. 2024;74(3): 229–263. https://doi.org/10.3322/caac.21834.

  2. Eastham JA, Auffenberg GB, Barocas DA, ChouR, Crispino T, Davis JW, et al. Clinically LocalizedProstate Cancer: AUA/ASTRO Guideline, Part II:Principles of Active Surveillance, Principles ofSurgery, and Follow-Up. The Journal of Urology.2022;208(1): 19–25. https://doi.org/10.1097/JU.0000000000002758.

  3. Lowrance WT, Breau RH, Chou R, Chapin BF,Crispino T, Dreicer R, et al. Advanced ProstateCancer: AUA/ASTRO/SUO Guideline PART I.Journal of Urology. 2021;205(1): 14–21. https://doi.org/10.1097/JU.0000000000001375.

  4. Tang C, Hoffman KE, Allen PK, Gabel M,Schreiber D, Choi S, et al. Contemporaryprostate cancer treatment choices inmultidisciplinary clinics referenced to nationaltrends. Cancer. 2020;126(3): 506–514. https://doi.org/10.1002/cncr.32570.

  5. Swindle P, Eastham JA, Ohori M, Kattan MW,Wheeler T, Maru N, et al. Do margins matter?The prognostic significance of positive surgicalmargins in radical prostatectomy specimens.The Journal of Urology. 2008;179(5 Suppl): S47-51. https://doi.org/10.1016/j.juro.2008.03.137.

  6. Yossepowitch O, Briganti A, Eastham JA,Epstein J, Graefen M, Montironi R, et al. PositiveSurgical Margins After Radical Prostatectomy: ASystematic Review and Contemporary Update.European Urology. 2014;65(2): 303–313.https://doi.org/10.1016/j.eururo.2013.07.039.

  7. John A, Lim A, Catterwell R, Selth L,O’Callaghan M. Length of positive surgicalmargins after radical prostatectomy: Doessize matter? - A systematic review and metaanalysis.Prostate Cancer and Prostatic Diseases.2023;26(4): 673–680. https://doi.org/10.1038/s41391-023-00654-6.

  8. Cartron E, Lecordier D, Jovic L. L’intelligenceartificielle en santé : une promesse, un défi etune opportunité de dialogue interdisciplinaire.Recherche en soins infirmiers. 2019;137(2): 5–5.https://doi.org/10.3917/rsi.137.0005.

  9. Hamet P, Tremblay J. Artificial intelligence inmedicine. Metabolism: Clinical and Experimental.2017;69S: S36–S40. https://doi.org/10.1016/j.metabol.2017.01.011.

  10. Van Booven DJ, Kuchakulla M, Pai R, Frech FS,Ramasahayam R, Reddy P, et al. A SystematicReview of Artificial Intelligence in ProstateCancer. Research and Reports in Urology.2021;13: 31–39. https://doi.org/10.2147/RRU.S268596.

  11. Chu TN, Wong EY, Ma R, Yang CH, DaliehIS, Hung AJ. Exploring the Use of ArtificialIntelligence in the Management of ProstateCancer. Current Urology Reports. 2023;24(5):231–240. https://doi.org/10.1007/s11934-023-01149-6.

  12. Denise Heady, MA. AI model may yield betteroutcomes for prostate cancer. UCLA Health.2024. https://www.uclahealth.org/news/release/ai-model-may-yield-better-outcomesprostate-cancer [Accessed 6th May 2026].

  13. Antonelli M, Johnston EW, Dikaios N, CheungKK, Sidhu HS, Appayya MB, et al. Machinelearning classifiers can predict Gleason pattern4 prostate cancer with greater accuracythan experienced radiologists. EuropeanRadiology. 2019;29(9): 4754–4764. https://doi.org/10.1007/s00330-019-06244-2.

  14. Radboud University Medical Center. AI betterdetects prostate cancer on MRI than radiologists.ScienceDaily. 2024. https://www.sciencedaily.com/releases/2024/06/240612113341.htm[Accessed 6th May 2026].

  15. Alitto AR, Gatta R, Vanneste B, Vallati M,Meldolesi E, Damiani A, et al. PRODIGE:PRediction models in prOstate cancer forpersonalized meDIcine challenGE. FutureOncology. 2017;13(24): 2171–2181. https://doi.org/10.2217/fon-2017-0142.

  16. National Comprehensive Cancer Network.Guidelines Detail. NCCN. https://www.nccn.org/guidelines/guidelines-detail?id=1459 [Accessed 6th May 2026].

  17. Lee K, Lee SH. Artificial Intelligence-DrivenOncology Clinical Decision SupportSystem for Multidisciplinary Teams. Sensors.2020;20(17): 4693. https://doi.org/10.3390/s20174693.

  18. Bostrom N. Superintelligence von NickBostrom - 978-0-19-166683-4. OxfordUniversity Press US; 2014. https://www.wirthmiller.at/item/Superintelligence/Nick_Bostrom?isbn=9780191666834 [Accessed6th May 2026].




2020     |     www.medigraphic.com