Observational Study Evaluate Pathology Practice Use Artificial Intelligence in Patient Suspected Lung and Breast Cancer (CASCADE)

A Non-interventional Study Evaluating Samples From Patients With Suspected Non-small Lung Cancer or Breast Cancer to Describe Pathology Practices and to Evaluate Computational Pathology Plus Artificial Intelligence Algorithms.

A multinational observational study to evaluate the current pathology practices and the utilization of computational pathology plus artificial intelligence algorithms in patients with suspected lung and breast cancer.

Study Overview

• Status: Completed
• Conditions: Breast Cancer; Lung Cancer

Detailed Description

A non-interventional study evaluating samples from patients with suspected non-small lung cancer or breast cancer to describe pathology practices and to evaluate computational pathology plus artificial intelligence algorithms in Australia, Brazil, Egypt, and Kenya. Use of digital and computational Artificial intelligence pathology in countries with low and high pathologist/population ratios is critical in developing a sustainable solution. The study has two parts, the first part will focus on breast cancer, and the second part will focus on lung cancer.

The laboratories have an active digital pathology setting and evaluate samples for cancer diagnosis. The centres of lung cancer part of the study will be selected at a later stage. The study will retrospectively evaluate samples from patients who have been preliminarily diagnosed with breast or lung cancer through clinical assessments and whose samples were evaluated only by using conventional workflow.

As part of the study, computational AI pathology algorithms will be implemented in each laboratory. Two AI pathology algorithms will be used in the breast cancer part of the study. Galen™ Breast application developed by Ibex Medical Analytics will be implemented in a laboratory in Australia. MindPeak Breast, developed by MindPeak GmbH will be implemented in laboratories in Brazil, Egypt, and Kenya. After implementing computational AI pathology algorithms, 150 samples evaluated for the primary objective from each laboratory for each cancer type will be evaluated using a conventional workflow plus an AI assisted workflow with human supervision and a conventional workflow plus an AI-assisted workflow without human supervision. These evaluations will be used to analyse secondary and exploratory objectives.

• Study Type: Observational
• Enrollment (Actual): 603

Contacts and Locations

Study Locations

• Brazil
  • São Paulo, Brazil
    • Research Site
• Kenya
  • Nairobi, Kenya
    • Research Site

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Child; Adult; Older Adult
• Accepts Healthy Volunteers: No

• Sampling Method: Probability Sample

Study Population

Data from samples that meet the following inclusion criterion will be analyzed.

• Sample from adult patients (≥ 18 years) with suspected non-small cell lung cancer or invasive breast cancer or ductal carcinoma in situ.

Description

Inclusion Criteria:Sample from adult patients (≥ 18 years) with suspected non-small cell lung cancer or invasive breast cancer or ductal carcinoma in situ.

-

Exclusion Criteria:

• Samples with the inadequate technical quality of slides (pre-analytics quality) or images, e.g., broken slides, large out-of-focus areas, slides with fixation artefacts.

  • Samples from cases that were included in the training or technical validation.
  • Sample taken by fine needle aspiration.
  • Sample sent for cytological evaluation.

Study Plan

How is the study designed?

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
primary objective	The duration between the biopsy-taken date/time and the biopsy-based pathological diagnosis date/time will be calculated based on the laboratory records retrospectively.	2 years
Primary Objective	Reading time to assess section slides for pathological diagnosis will also be extracted from the laboratory records, if relevant information was kept in the records.	2 years
exploratory objective	the total cost and fees related to training, for implementing digital pathology and computational AI pathology algorithms will be assessed as an endpoint.	2 years

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
secondary objectives	agreement rate :PPV and NPV for computational AI pathology algorithms (with and without human supervision) when the conventional pathology workflow is the reference will also be evaluated.	2 years
exploratory objective	the total cost and fees related to employees for implementing digital pathology and computational AI pathology algorithms will be assessed as an endpoint.	2 years

Other Outcome Measures

Outcome Measure	Measure Description	Time Frame
Secondary Objective	sensitivity:PPV and NPV for computational AI pathology algorithms (with and without human supervision) when the conventional pathology workflow is the reference will also be evaluated.	2 years
Secondary Objective	specificity: PPV and NPV for computational AI pathology algorithms (with and without human supervision) when the conventional pathology workflow is the reference will also be evaluated.	2 years
exploratory objective	The total cost and fees related to hardware for implementing digital pathology and computational AI pathology algorithms will be assessed as an endpoint.	2 years
Exploratory objective	the total cost and fees related to software for implementing digital pathology and computational AI pathology algorithms will be assessed as an endpoint.	2 years

Collaborators and Investigators

• Sponsor: AstraZeneca

Study record dates

Study Major Dates

• Study Start (Actual): October 25, 2025
• Primary Completion (Actual): March 19, 2026
• Study Completion (Actual): March 19, 2026

Study Registration Dates

• First Submitted: January 28, 2025
• First Submitted That Met QC Criteria: February 13, 2025
• First Posted (Actual): February 14, 2025

Study Record Updates

• Last Update Posted (Actual): September 3, 2026
• Last Update Submitted That Met QC Criteria: September 2, 2026
• Last Verified: August 1, 2026

More Information

Terms related to this study

• Keywords: Artificial intelligence; Algorithm; Positive predictive value; Negative predictive value; Computational Pathology; Galen™ Breast application; MindPeak GmBH; Laboratories
• Additional Relevant MeSH Terms: Neoplasms by Site; Neoplasms; Respiratory Tract Diseases; Lung Diseases; Respiratory Tract Neoplasms; Thoracic Neoplasms; Skin Diseases; Breast Diseases; Skin and Connective Tissue Diseases; Lung Neoplasms; Breast Neoplasms
• Other Study ID Numbers: D4191R00089

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: YES
• IPD Plan Description: Qualified researchers can request access to anonymized individual patient-level data from AstraZeneca group of companies sponsored clinical trials via the request portal Vivli.org. All requests will be evaluated as per the AZ disclosure commitment: https://astrazenecagrouptrials.pharmacm.com/ST/Submission/Disclosure. Yes, indicates that AZ are accepting requests for IPD, but this does not mean all requests will be shared.
• IPD Sharing Time Frame: AstraZeneca will meet or exceed data availability as per the commitments made to the EFPIA PhRMA Data Sharing Principles. For details of our timelines, please rerefer to our disclosure commitment at https://astrazenecagrouptrials.pharmacm.com/ST/Submission/Disclosure.

IPD Sharing Access Criteria

When a request has been approved AstraZeneca will provide access to the anonymized individual patient-level data via secure research environment Vivli.org.

Signed Data Usage Agreement (non-negotiable contract for data accessors) must be in place before accessing requested information.

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No