Using artificial intelligence to help doctors and scientists in the lab that studies diseases in the body, in order to use less of a specific testing method called immunohistochemistry and make the work process smoother and more efficient.

ISRCTN	ISRCTN14323711
DOI	https://doi.org/10.1186/ISRCTN14323711

Submission date: 22/05/2023
Registration date: 31/05/2023
Last edited: 26/05/2023

Recruitment status: No longer recruiting
Overall study status: Completed
Condition category: Other

Prospectively registered

Protocol

Statistical analysis plan

Results

Individual participant data

Record updated in last year

Plain English summary of protocol

Background and study aims
In this single-center trial, we aim to investigate an artificial intelligence (AI)-assisted workflow. We investigate whether an AI-assisted workflow reduces expensive immunohistochemistry stains for the detection of cancer cells in lymph nodes in breast cancer patients, and in prostate needle biopsies in patients suspected of prostate cancer. Meanwhile, we maintain current diagnostic safety standards.

Who can participate?
All patients at UMC Utrecht are automatically included

What does the study involve?
Patients' material will be assessed by pathologists conform the 'standard-of-care'-workflow, or in an AI-assisted workflow. Every two weeks, pathologists will (prospectively) assess pathology specimens either with or without AI-assistance.

What are the possible benefits and risks of participating?
Neither benefits or risks are applicable to patients, as we have a diagnostic safety net in place. If no tumor can be seen, we will perform the additional immunohistochemistry, so no tumors will be missed.

Where is the study run from
UMC Utrecht (Netherlands)

When is the study starting and how long is it expected to run for?
January 2022 to December 2023

Who is funding the study?
The Hanarth fund (Netherlands)
Paige (USA)

Who is the main contact?
Paul J van Diest, p.j.vandiest@umcutrecht.nl

Contact information

Prof Paul van Diest
Principal Investigator

Department of Pathology, PO Box 85500, 3508 GA Utrecht, The Netherlands
Utrecht
3508 GA
Netherlands

ORCID ID	0000-0003-0658-2745
Phone	+31887556565
Email	p.j.vandiest@umcutrecht.nl

Study information

Study design	Single-center interventional non-randomized trial
Primary study design	Interventional
Secondary study design	Non randomised study
Study setting(s)	Laboratory
Study type	Diagnostic, Safety, Efficacy
Participant information sheet	No participant information sheet available
Scientific title	Clinical implementation of artificial intelligence assistance in pathology workflow
Study acronym	CONFIDENT
Study objectives	The number of required immunohistochemistry (IHC) stains used to confirm diagnosis will be reduced in the AI-assisted arm, compared to the standard-of-care arm.
Ethics approval(s)	The ethics committee (MREC NedMec) waived the need of official ethical approval, since participants are not subjected to procedures nor are they required to follow rules.
Health condition(s) or problem(s) studied	Reduction of immunohistochemistry used for the diagnosis of lymph node metastases in breast cancer (BCa - CONFIDENT-B), and prostate cancer (PCa - CONFIDENT-P) in prostate needle biopsies.
Intervention	During the study period, all whole slide images (WSI) will be assessed by the same group of pathologists; i.e. two expert urological pathologists for the prostate needle biopsies, and three expert breast pathologists for the lymph node assessment from BCa patients. For both the CONFIDENT-B and CONFIDENT-P trial, the specific pathology specimens will be assigned to be assessed by a pathologist with or without AI-assistance in a pragmatic (bi-)weekly sequential design. For obvious reasons, allocation concealment and blinding of pathologists and researchers is not applicable. All eligible specimens of patients will be assigned to either the control group or the intervention group. In the control group, pathologists will assess haematoxylin eosin (HE)-stained WSI of patients digitally, according to the current clinical workflow. For prostate biopsies, IHC is routinely performed on all cases. For BCa lymph nodes, if no metastases or tumour are present, IHC staining will be performed. Additional IHC staining will also be performed by additional request of the pathologist in case of doubt. In the intervention group, pathologists will assess the HE-specimens digitally with the outcome of the algorithm provided in their first assessment of the specimen. For PCa, they will use the CE-IVD certified Paige Prostate Suite algorithms for tumor detection and tumor volume percentage calculations. For BCa, pathologists will use the CE-IVD certified Metastasis Detection App by Visiopharm. AI analysis of the WSI will be performed right after scanning to avoid delays in the clinical workflow. If the AI-assisted pathologist does not detect metastases or tumours on the HE slide, routine additional IHC staining will be performed by P503S/p63/CK HMW for PCa and CAM5.2 for BCa, to ensure no metastases or tumours are missed. Pathologists can always request an additional IHC-stain if they feel they need this to make an adequate diagnosis.
Intervention type	Device
Pharmaceutical study type(s)	Not Applicable
Phase	Not Applicable
Drug / device / biological / vaccine name(s)	Paige Prostate Suite and Metastasis Detection App by Visiopharm
Primary outcome measure	Relative risk of IHC-use per detected case of SN-metastases and risk of IHC-use per detected tumor in prostate needle biopies (case- and slide-level) measured using number of spent resources, i.e. the number of IHC-stains performed in both groups, measured after the single visit
Secondary outcome measures	1. Sensitivity and negative-predictive value of the AI-assisted pathologist. We use the assessment of the pathologist on immunohistochemistry stained slides as a reference standard. These slides are always performed in cases where the pathologist is in doubt, or where they think the slides are benign. Obvious malignant cases are spared an IHC staining (as our hypothesis is that these obvious malignant cases will be higher in the AI-assisted arm than in the standard of care-arm). 2. Differences in mean reading time per H&E-slide and per case between study arms, measured during the primary WSI assessment. 3. The number of IHC stains that may have been omitted after AI-implementation by analyzing standalone performances of the AI on the standard-of-care arm retrospectively 4. Pathologists’ evaluation by a questionnaire on the AI-assisted work process after ending of the study enrollment. 5. Stand-alone performance of the algorithm. We measure sensitivity and negative-predictive value using the pathologists’ assessment as a ground truth (with or without IHC, as specified above), performed retrospectively on the control-arm as well as on the AI-assisted arm. Specificity and positive predictive value cannot be calculated because of the study design and are therefore not incorporated in the stand-alone performance nor in the AI-assisted pathologists’ performance. 6. Difference in diagnostic confidence of the pathologists between the study arms, measured on a 5-Likert scale during the primary WSI assessment. Sensitivity and specificity analyses of the algorithm itself have already been well documented, and is therefore outside the scope of the paper, as we focus on the combination of pathologist and AI to explore cost savings.
Overall study start date	01/01/2022
Completion date	31/12/2023

Eligibility

Participant type(s)	Patient, Health professional
Age group	Adult
Sex	Both
Target number of participants	PCa: 80 BCa: 180
Key inclusion criteria	Health professionals: 1.1. PCa: dedicated uropathologists 1.2. BCa: dedicated breast pathologists Patients: 1.1. PCa: adult males of any age undergoing prostate needle biopsies at the investigation site will be enrolled in the study and have their diagnosis determined in a prospective, consecutive manner. 1.2. BCa: adult females or males of any age with breast cancer as primary malignancy whose SN-specimen is assessed at the investigation site will be enrolled in the study and have their diagnosis determined in a prospective, consecutive manner. 2. All specimens that are H&E-stained and glass or film cover-slipped will be evaluated. Adjacent unstained slides will available for IHC staining. 3. The WSIs fulfill the quality checks described in the scanner manufacturer’s Instruction for Use and general clinical practice. 4. No diagnostic markings and patient identifiable markings are visible on the slides.
Key exclusion criteria	1. Patients who were referred for a second opinion 2. Patients that have opted out of all medical research 3. Cases in which the associated IHC-stained slides are unavailable/cannot be generated 4. Cases not meeting defined quality criteria for digital clinical primary diagnosis evaluation
Date of first enrolment	19/09/2022
Date of final enrolment	31/12/2023

Locations

Countries of recruitment

Netherlands

Study participating centre

University Medical Center Utrecht

Department of Pathology
PO Box 85500
Utrecht
3508 GA
Netherlands

Sponsor information

University Medical Center Utrecht
Hospital/treatment centre

Heidelberglaan 100
Utrecht
3584 CX
Netherlands

Phone	+31 887556565
Email	R.A.W.vanLier-2@umcutrecht.nl
Website	http://www.umcutrecht.nl/nl/
"ROR"	https://ror.org/0575yy874

Funders

Funder type

Not defined

Hanarth Fonds
Private sector organisation / Trusts, charities, foundations (both public and private)

Alternative name(s): Hanarth Foundation, Hanarth Fund, Hanarth Funds Foundation, Stichting Hanarth Fonds
Location: Netherlands

Paige

No information available

Results and Publications

Intention to publish date	31/12/2024
Individual participant data (IPD) Intention to share	No
IPD sharing plan summary	Data sharing statement to be made available at a later date
Publication and dissemination plan	Planned publication in a peer-reviewed journal.
IPD sharing plan	The data-sharing plans for the current study are unknown and will be made available at a later date

Editorial Notes

26/05/2023: Trial's existence confirmed by MREC NedMec