An international multi-institutional validation study of the algorithm for prostate cancer detection and Gleason grading

npj Precision Oncology, vol. 7, pp. 77

Abstract

Pathologic examination of prostate biopsies is time consuming due to the large number of slides per case. In this retrospective study, we validate a deep learning-based classifier for prostate cancer (PCA) detection and Gleason grading (AI tool) in biopsy samples. Five external cohorts of patients with multifocal prostate biopsy were analyzed from high-volume pathology institutes. A total of 5922 H&E sections representing 7473 biopsy cores from 423 patient cases (digitized using three scanners) were assessed concerning tumor detection. Two tumor-bearing datasets (core n = 227 and 159) were graded by an international group of pathologists including expert urologic pathologists (n = 11) to validate the Gleason grading classifier. The sensitivity, specificity, and NPV for the detection of tumor-bearing biopsies was in a range of 0.971-1.000, 0.875-0.976, and 0.988-1.000, respectively, across the different test cohorts. In several biopsy slides tumor tissue was correctly detected by the AI tool that was initially missed by pathologists. Most false positive misclassifications represented lesions suspicious for carcinoma or cancer mimickers. The quadratically weighted kappa levels for Gleason grading agreement for single pathologists was 0.62-0.80 (0.77 for AI tool) and 0.64-0.76 (0.72 for AI tool) for the two grading datasets, respectively. In cases where consensus for grading was reached among pathologists, kappa levels for AI tool were 0.903 and 0.855. The PCA detection classifier showed high accuracy for PCA detection in biopsy cases during external validation, independent of the institute and scanner used. High levels of agreement for Gleason grading were indistinguishable between experienced genitourinary pathologists and the AI tool.

Authors 22

1. Yuri Tolkach corresponding

   University Hospital Cologne

   Affiliation as printed

   Institute of Pathology, University Hospital Cologne, Cologne, Germany. yuri.tolkach@gmail.com

   Institute of Pathology, University Hospital Cologne, Cologne, Germany

2. Vlado Ovtcharov
   Affiliation as printed

   Indica Labs, Albuquerque, NM, USA

3. Alexey E. Pryalukhin

   Universitätsklinikum Wiener Neustadt

   Affiliation as printed

   Institute of Pathology, Landesklinikum Wiener Neustadt, Wiener Neustadt, Austria

4. Marie‐Lisa Eich

   University Hospital Cologne

   Affiliation as printed

   Institute of Pathology, University Hospital Cologne, Cologne, Germany

5. Nadine Therese Gaisa Aachen Institute of Pathology

   Universitätsklinikum Aachen

   Affiliation as printed

   Institute of Pathology, University Hospital Aachen, Aachen, Germany

6. MARTIN A. BRAUN
   Affiliation as printed

   Institute of Pathology Troisdorf, Troisdorf, Germany

7. Abdukhamid Radzhabov
   Affiliation as printed

   Urology Clinic, Municipal Clinic of Brunswick, Brunswick, Germany

8. Alexander Quaas

   University Hospital Cologne

   Affiliation as printed

   Institute of Pathology, University Hospital Cologne, Cologne, Germany

9. Peter Hammerer
   Affiliation as printed

   Urology Clinic, Municipal Clinic of Brunswick, Brunswick, Germany

10. Ansgar Dellmann
    Affiliation as printed

    Institute of Pathology, Municipal Clinic of Brunswick, Brunswick, Germany

11. Wolfgang Hulla

    Universitätsklinikum Wiener Neustadt

    Affiliation as printed

    Institute of Pathology, Landesklinikum Wiener Neustadt, Wiener Neustadt, Austria

12. Michael C. Haffner

    Fred Hutch Cancer Center

    Affiliation as printed

    Divisions of Human Biology and Clinical Research, Fred Hutch Cancer Center, Seattle, WA, USA

13. Henning Reis

    Goethe University Frankfurt · University Hospital Frankfurt

    Affiliation as printed

    Dr. Senckenberg Institute of Pathology, University Hospital Frankfurt, Goethe University Frankfurt, Frankfurt am Main, Germany

14. Ibrahim Fahoum

    Tel Aviv Sourasky Medical Center

    Affiliation as printed

    Department of Pathology, Sourasky Medical Center, Tel Aviv, Israel

15. Iryna V. Samarska

    Maastricht University Medical Centre

    Affiliation as printed

    Department of Pathology, University Hospital Maastricht, Maastricht, The Netherlands

16. A.M. Borbat

    State Scientific Center of the Russian Federation - Federal Medical Biophysical Center named after A.I. Burnazyan · Federal Medical-Biological Agency

    Affiliation as printed

    Department of Pathology, Burnasyan Federal Medical Biophysical Center of Federal Medical Biological Agency, Moscow, Russia

17. Hoa Hoang Ngoc Pham

    University of Nagasaki · Bạch Mai Hospital · Nagasaki University

    Affiliation as printed

    Department of Pathology, Bach Mai Hospital, Hanoi, Vietnam

    Department of Pathology, University of Nagasaki, Nagasaki, Japan

18. Axel Heidenreich

    University Hospital Cologne

    Affiliation as printed

    Clinic of Urology, University Hospital Cologne, Cologne, Germany

19. Sebastian Klein

    University Hospital Cologne

    Affiliation as printed

    Institute of Pathology, University Hospital Cologne, Cologne, Germany

20. George Jabboure Netto

    University of Alabama · University of Alabama at Birmingham

    Affiliation as printed

    Department of Pathology, University of Alabama, Birmingham, AL, USA

21. Peter D. Caie
    Affiliation as printed

    Indica Labs, Albuquerque, NM, USA

22. Reinhard Buettner

    University Hospital Cologne

    Affiliation as printed

    Institute of Pathology, University Hospital Cologne, Cologne, Germany