Table 2.
Summary of major digital pathology validation studies which include urological specimens
| AUTHOR | SCANNING SYSTEM EVALUATED | SCANNING MAGNIFICATION | STUDY METHODOLOGY | UROLOGICAL CASES INCLUDED IN STUDY | AREAS OF DISCORDANCE BETWEEN GLASS SLIDE (GS) READ AND DIGITAL PATHOLOGY (DP) READ FOR PURPOSE OF STUDY (EXCLUDING ORIGINAL DIAGNOSIS OR ADJUDICATED DIAGNOSIS WHERE GIVEN) |
|---|---|---|---|---|---|
| Campbell 2012 [5] | iScan® | x20 |
Single centre study 2 pathologists Digital diagnosis compared with original diagnoses (no washout period as the original glass slide diagnosis was the comparator), consensus diagnosis for discrepant cases |
N = 6 Type not specified |
Prostate biopsy (1 case) – benign (GS) vs. ASAP (DP) Bladder specimen (type not specified, 1 case) – suggestive of polypoid cystitis (GS) vs. PUNLMP (DP) |
| Bauer 2013 [3] | Aperio®ScanScope XT | x20 |
Multicentre study 2 pathologists Glass diagnosis followed by washout period of 1 year, then either digital or GS review, consensus diagnosis for discrepant cases |
Type not specified | Prostate specimens (type not specified, 6 cases) – Gleason grading (x3 cases), benign (GS) vs. HG PIN (DP), PIN (GS) vs. benign (DP) |
| Al-Janabi 2014 [19] | Not specified | x20 |
Single centre study 2 pathologists Digital diagnosis compared with original diagnoses (each pathologist reviewed only the GS cases they originally diagnosed), washout period minimum 6 months. The original glass slide diagnosis was the comparator, consensus diagnosis for discrepant cases |
N = 100 Urinary system only Kidney = 50 (48 cases were medical renal disease diagnoses, 2 surgical diagnoses) Bladder = 43 Ureter = 1 Urethra = 6 |
Excluding the medical renal cases Bladder specimens (5 cases) – Grade 3 PUC, non-invasive (GS) vs. grade 3 PUC with lamina propria invasion (DP), benign no abnormality (GS) vs. chronic inflammation (DP), grade 3 PUC, suspicious for invasion (GS) vs. grade 3 PUC with lamina propria invasion and CIS (DP), & 2 cases with minor descriptive discrepancy only |
| Snead 2016 [8] | Omnyx® VL4 | x40 |
Single centre study 17 pathologists Glass slide diagnosis, followed by washout period of 21 days, before digital read (33 % cases reported on GS and digital by the same pathologist, and 66 % by two different pathologists), consensus diagnosis for discrepant cases |
N = 242 Type not specified |
Penile biopsy with HPV changes and atypia (GS) vs. PeIN (DP) Prostate biopsies (2 cases) – Gleason grading (pattern 4 vs. 3), suspicious for malignancy (GS) vs. benign (DP) Urothelial biopsies (3 cases) – Urothelial carcinoma grade 1 LG (GS) vs. grade 2 HG (DP), urothelial carcinoma with no CIS (GS) vs. urothelial carcinoma with CIS (DP), non-invasive urothelial carcinoma (GS) vs. urothelial carcinoma with early invasion (DP) |
| Tabata 2017 [9] |
PhilipsIntelliSite® Ultrafast scanner Leica Biosystems® Aperio ®AT2 scanner Hamamatsu® Nanozoomer® 2.0-HT C9600-13 Hamamatsu® NanoZoomer® 2.0-RS C10730-13 CLARO FINO |
Variable, x20 and x40 |
Multicentre study 9 pathologists Each pathologist carried out digital and glass slide reads on each case, washout period minimum of 14 days, consensus diagnosis for discrepant cases |
N = 66 |
Prostate specimen (type not specified, 1 case) – benign (GS) vs. atypical glands (DP) Genitourinary organ (not specified, 1 case) – erosive mucosa without malignancy (GS) vs. erosive mucosa possible for malignancy (DP) |
| Mukhopadhyay 2018 [7] | Philips IntelliSite® Pathology Solution | Not specified |
Multicentre study 16 pathologists Each pathologist carried out digital and glass slide reads on each case, washout period minimum of 16 days, reference standard = original GS diagnosis, with adjudication for discrepant cases |
N = 448 Urinary bladder = 99 Prostate = 299 Kidney, neoplastic = 50 |
Kidney, neoplastic (2 cases) – papillary RCC (GS) vs. metanephric adenoma (DP), HG papillary urothelial carcinoma (GS) vs. LG urothelial carcinoma (DP) Urinary bladder (not otherwise specified) (18 cases) – cystitis with reactive atypia (GS) vs. HG CIS (DP), HG non-invasive PUC (GS) vs. HG PUC with lamina propria invasion (DP), HG PUC without invasion (GS) vs. LG PUC non-invasive (DP), LG PUC non-invasive (GS) vs. LG PUC with invasion (DP), cystitis with hyperplasia (GS) vs. flat HG dysplasia (DP), HG urothelial carcinoma (GS) vs. fibrosis, benign (DP), tissue highly suspicious for invasive SCC (GS) vs. mildly atypical squamous epithelium, favour squamous metaplasia (DP), HG urothelial carcinoma invading through bladder wall into perivesical soft tissue (GS) vs. benign (DP), cystitis with mucosal ulceration with reactive atypia (GS) vs. myoinvasive HG urothelial carcinoma (DP), cystitis with granulomatous features and reactive atypia (GS) vs. myoinvasive HG urothelial carcinoma (DP), CIS (GS) vs. cystitis (DP), HG PUC non-invasive (GS) vs. HG PUC with lamina propria invasion (DP), cystitis with reactive atypia (GS) vs. CIS (DP), CIS (GS) vs. HG urothelial carcinoma with lamina propria invasion (DP), atypical urothelium (GS) vs. benign (DP), CIS (GS) vs. inflammation (DP), CIS (GS) vs. inflammation (DP), CIS (GS) vs. inflammation (DP) |
| Vodovnik 2018 [13] | Aperio® ScanScope® AT Turbo | x20 |
Single centre study 1 pathologist Digital diagnosis compared with original diagnosis (the pathologist had reported the cases on both GS and DP), washout period 6 months. |
N = 75 |
Prostatic adenocarcinoma grading Gleason 9 (5 + 4) (GS) vs. Gleason 9 (4 + 5) (DP) UC HG grade 2–3 (GS) vs. UC HG grade 2 (DP) |
| Borowsky 2020 [4] | Leica Biosystems® Aperio® AT2 DX system | x20 |
Multicentre study 19 pathologists Each pathologist carried out digital and glass slide reads on each case, washout period minimum of 31 days, reference standard = original GS diagnosis, with adjudication for discrepant cases |
N = 447 Urinary bladder = 100 Prostate = 300 Kidney, neoplastic = 47 |
Not specified, although comment that urinary bladder biopsies showed the highest major discrepancy rate |
| Hanna 2020 [11] | Leica Biosystems® Aperio® GT450 | x40 |
Single centre study 12 pathologists (2 reporting GU cases) Each pathologist carried out digital and glass slide reads on each case, the digital read was done remotely via a virtual private network (VPN), and the GS read was done on site in the hospital department with a mean interval of 2 days. Reference standard = GS diagnosis with adjudication for discordant cases. |
718 slides = 108 cases in total across specialties, and for GU the following specimens; Prostate = 151 Bladder = 28 Lymph nodes = 10 Kidney = 9 Urethra = 6 Testis = 3 Ureter = 1 Adrenal = 1 Other = 10 |
No major or minor discordances |
| Rao 2021 [12] | Ventana® DP200 | x20 (x40 scanning available on request) |
Single centre study 18 pathologists Study looked at concordance between digital sign out of cases remotely (from home) with blinded re-review of cases after a minimum 2 week interval. Concordance adjudicated by a referee pathologist not participating in the sign out study. Blind consensus diagnosis established for discordant diagnoses. |
N = 25 (1 of which was deferred to glass) 47 parts, 74 slides; Urinary bladder, ureteric orifice = 24 Kidney = 5 Penis = 2 Prostate = 12 Iliac fossa = 1 Lung = 1 Endometrium = 1 Rectum = 1 |
Urinary bladder (1 case, TURBT) HG PUC T1 (GS) vs. HG PUC Ta (DP) |
GS glass slides, DP digital pathology, ASAP atypical small acinar proliferation, PUNLMP papillary urothelial neoplasm of low malignant potential, PIN prostatic intraepithelial neoplasia, PUC papillary urothelial carcinoma, CIS carcinoma in situ, HPV human papilloma virus, PeIN penile intraepithelial neoplasia, LG WHO 2004 low grade, HG WHO 2004 high grade, RCC renal cell carcinoma, SCC squamous cell carcinoma, GU genitourinary, TURBT transurethral resection of bladder tumour