Colon Adenoma, Colon Lesion, Colon Polyp
Conditions
Brief summary
This is a prospective, multicenter, randomized controlled study to evaluate the effect of the Computer-Assisted Detection (CADe) Device on Adenomas Per Colonoscopy and Positive Percent Agreement for routine colonoscopies. The control arm is colonoscopy performed with High Definition White Light Endoscopy (HD-WLE) per standard of care. The intervention arm is colonoscopy performed with HD-WLE per standard of care plus the Computer-Assisted Detection (CADe) Device.
Interventions
DEVICEComputer-Assisted Detection (CADe) Device
The CADe Device uses artificial intelligence to aid in identifying colorectal polyps during High Definition White Light Endoscopy (HD-WLE) based colonoscopies. This device connects the Gastroenterologist's colonoscopy video output source to the main monitor and highlights the regions of interest where the device detects a potential lesion. The CADe Device overlays graphical markers onto video from the endoscope camera and does not perform further processing to the endoscope video signal. This device is not intended to replace clinical decision making.
Sponsors
Verily Life Sciences LLC
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
SCREENING
Masking
SINGLE (Outcomes Assessor)
Masking description
Subjects, investigators, study site personnel, and pathologists will not be blinded to treatment assignments. All Verily personnel involved in the statistical analysis of this study will be blinded to treatment assignment. For the interim analysis, unblinded analysis will be performed by an independent statistician, the results of which will be communicated to a select group of Verily personnel not involved in the study management.
Eligibility
Sex/Gender
ALL
Age
45 Years to 80 Years
Healthy volunteers
Yes
Inclusion criteria
* Scheduled to undergo routine screening (including, but not limited to, FIT/Cologuard positive), routine surveillance (≥3 years as scheduled since last colonoscopy), or diagnostic (symptomatic) colonoscopy with High Definition White Light Endoscopy. * Between the ages of 45 and 80 years, inclusive * Able and willing to provide written informed consent
Exclusion criteria
* Self-reported pregnancy * Known diagnosis of Colorectal Cancer * History of, or referral for, Inflammatory Bowel Disease * Previous surgery involving the colon or rectum * Referral for known polyp or assessment of post-polypectomy site (i.e. less than 3 years since last colonoscopy). * High suspicion or diagnosis of genetic polyposis syndromes, including familial adenomatous polyposis (FAP), hereditary nonpolyposis colorectal cancer (HNPCC), or any other high-risk family history meeting Bethesda guidelines. * Referral for overt, symptomatic gastrointestinal bleeding
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Difference in Adenomas Per Colonoscopy (APC) | Day 1 | Difference in Adenomas Per Colonoscopy (APC) between the control and intervention arm, evaluated for superiority. APC is defined as the average number of histologically confirmed adenomas resected per colonoscopy. |
| Difference in Positive Percent Agreement (PPA) | Day 1 | Difference in Positive Percent Agreement (PPA) between the control and intervention arm, evaluated for non-inferiority. PPA is defined as the total number of histologically confirmed Clinically Significant Excised Lesions, divided by the total number of excisions. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Mean Withdrawal and Inspection Time (MWT) | Day 1 | The withdrawal time is defined as the time measured from the moment the withdrawal phase of the procedure begins (with the scope in the cecum) to the moment the scope is withdrawn from the patient. The Inspection time measurement will exclude washing and resection, and other peri-resection activity not deemed to be colonic inspection. Inspection times for both the control arm and intervention arm will be calculated retrospectively upon review of the video recordings. |
| Polyp Detection Rate (PDR) | Day 1 | Polyp detection rate is defined as the proportion of patients with at least one histologically-confirmed polyp detected. |
| Proximal Adenoma Detection Rate (pADR) | Day 1 | pADR is defined as the percentage of patients with at least one histologically-confirmed adenoma detected in proximal colon. |
| Flat Adenoma Detection Rate (fADR) | Day 1 | fADR is defined as the percentage of patients with at least one histologically-confirmed non-polypoid adenoma detected. |
| Serrated Lesions Per Colonoscopy (SLPC) | Day 1 | SLPC is defined as the number of histologically confirmed serrated lesions detected, divided by the total number of colonoscopies. |
| Adenoma Detection Rate (ADR) | Day 1 | The Adenoma Detection Rate is defined as the number of patients with at least one histologically confirmed adenoma divided by the total number of patients enrolled per study arm. |
| Adenoma Detection Rate Including Carcinoma (ADR*) | Day 1 | ADR\* is defined as ADR, but also includes histologically-confirmed intramucosal carcinoma and adenocarcinoma. |
| Small Adenoma Detection Rate (sADR) | Day 1 | sADR is defined as the percentage of patients with at least one adenoma 5mm or smaller detected. |
| Polyps Per Colonoscopy (PPC) | Day 1 | PPC is defined as the total number of histologically-confirmed polyps found divided by the total number of colonoscopies performed, per study arm. |
| Advanced Adenoma Detection Rate (aADR) | Day 1 | aADR is defined as the percentage of patients with at least one adenoma ≥ 10 mm, or any adenoma \< 10 mm, which was either of high-grade dysplasia (HGD) or villous or tubulovillous. |
| False Positive Rate (FPR) | Day 1 | FPR is defined as the proportion of colorectal lesions resected and biopsied and subsequently not histologically-confirmed to be clinically relevant colorectal polyps (e.g. a pathology finding of normal mucosa, inflammatory tissue, stool or debris, lymphoid aggregates). |
| Serrated Lesions Detection Rate (SLDR) | Day 1 | SLDR is defined as the percentage of patients with at least one histologically confirmed serrated lesion detected. |
| Number of False Alerts Per Procedure | Day 1 | A false alert is defined as a bounding box that persists on the screen (approximately 2-3 seconds per the judgment of the colonoscopist) that is then determined by the colonoscopist not to contain a polyp. The false alert rate is calculated as the number of false alerts per procedure conducted in the intervention arm of the study. |
Countries
Israel, United States
Participant flow
Participants by arm
| Arm | Count |
|---|---|
| Colonoscopy (Standard of Care) The control arm is colonoscopy with High Definition White Light Endoscopy (HD-WLE) per standard of care. | 694 |
| CADe Device The intervention arm is colonoscopy with HD-WLE per standard of care plus the CADe Device.
Computer-Assisted Detection (CADe) Device: The CADe Device uses artificial intelligence to aid in identifying colorectal polyps during High Definition White Light Endoscopy (HD-WLE) based colonoscopies. This device connects the Gastroenterologist's colonoscopy video output source to the main monitor and highlights the regions of interest where the device detects a potential lesion. The CADe Device overlays graphical markers onto video from the endoscope camera and does not perform further processing to the endoscope video signal. This device is not intended to replace clinical decision making. | 713 |
| Total | 1,407 |
Baseline characteristics
| Characteristic | Total | Colonoscopy (Standard of Care) | CADe Device |
|---|---|---|---|
| Age, Continuous | 60.54 years STANDARD_DEVIATION 9.05 | 60.42 years STANDARD_DEVIATION 8.94 | 60.66 years STANDARD_DEVIATION 9.16 |
| Colonoscopy Indication Diagnostic (symptomatic) | 99 Participants | 44 Participants | 55 Participants |
| Colonoscopy Indication Screening | 750 Participants | 373 Participants | 377 Participants |
| Colonoscopy Indication Surveillance | 558 Participants | 277 Participants | 281 Participants |
| Race (NIH/OMB) American Indian or Alaska Native | 3 Participants | 2 Participants | 1 Participants |
| Race (NIH/OMB) Asian | 17 Participants | 7 Participants | 10 Participants |
| Race (NIH/OMB) Black or African American | 48 Participants | 22 Participants | 26 Participants |
| Race (NIH/OMB) More than one race | 2 Participants | 1 Participants | 1 Participants |
| Race (NIH/OMB) Native Hawaiian or Other Pacific Islander | 5 Participants | 1 Participants | 4 Participants |
| Race (NIH/OMB) Unknown or Not Reported | 14 Participants | 9 Participants | 5 Participants |
| Race (NIH/OMB) White | 1318 Participants | 652 Participants | 666 Participants |
| Region of Enrollment Israel | 295 participants | 143 participants | 152 participants |
| Region of Enrollment United States | 1112 participants | 551 participants | 561 participants |
| Sex: Female, Male Female | 752 Participants | 369 Participants | 383 Participants |
| Sex: Female, Male Male | 655 Participants | 325 Participants | 330 Participants |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk |
|---|---|---|
| deaths Total, all-cause mortality | 0 / 694 | 0 / 713 |
| other Total, other adverse events | 0 / 694 | 0 / 713 |
| serious Total, serious adverse events | 0 / 694 | 0 / 713 |
Outcome results
Primary
Difference in Adenomas Per Colonoscopy (APC)
Difference in Adenomas Per Colonoscopy (APC) between the control and intervention arm, evaluated for superiority. APC is defined as the average number of histologically confirmed adenomas resected per colonoscopy.
Time frame: Day 1
Population: Intent-to-Treat
| Arm | Measure | Value (LEAST_SQUARES_MEAN) |
|---|---|---|
| Colonoscopy (Standard of Care) | Difference in Adenomas Per Colonoscopy (APC) | 0.60 Resected adenomas |
| CADe Device | Difference in Adenomas Per Colonoscopy (APC) | 0.74 Resected adenomas |
p-value: 0.000295% CI: [0.07, 0.23]Poisson Regression
Primary
Difference in Positive Percent Agreement (PPA)
Difference in Positive Percent Agreement (PPA) between the control and intervention arm, evaluated for non-inferiority. PPA is defined as the total number of histologically confirmed Clinically Significant Excised Lesions, divided by the total number of excisions.
Time frame: Day 1
Population: Intent-to-Treat
| Arm | Measure | Value (LEAST_SQUARES_MEAN) |
|---|---|---|
| Colonoscopy (Standard of Care) | Difference in Positive Percent Agreement (PPA) | 0.62 Proportion of significant lesions |
| CADe Device | Difference in Positive Percent Agreement (PPA) | 0.56 Proportion of significant lesions |
p-value: 0.0995% CI: [-0.15, 0.03]Resampling based
Secondary
Adenoma Detection Rate (ADR)
The Adenoma Detection Rate is defined as the number of patients with at least one histologically confirmed adenoma divided by the total number of patients enrolled per study arm.
Time frame: Day 1
Population: Intent-to-Treat
| Arm | Measure | Value (NUMBER) |
|---|---|---|
| Colonoscopy (Standard of Care) | Adenoma Detection Rate (ADR) | 38.9 Percentage of participants with adenomas |
| CADe Device | Adenoma Detection Rate (ADR) | 42.9 Percentage of participants with adenomas |
p-value: 0.03195% CI: [-0.3, 11.2]Cochran-Mantel-Haenszel
Secondary
Adenoma Detection Rate Including Carcinoma (ADR*)
ADR\* is defined as ADR, but also includes histologically-confirmed intramucosal carcinoma and adenocarcinoma.
Time frame: Day 1
Population: Intent-to-Treat
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Colonoscopy (Standard of Care) | Adenoma Detection Rate Including Carcinoma (ADR*) | 271 Participants |
| CADe Device | Adenoma Detection Rate Including Carcinoma (ADR*) | 307 Participants |
Comparison: Difference in percentage between both arms is reported. H0: Colonoscopy (Standard of Care) percentage \<= CADe Device percentage (i.e. test of superiority)p-value: 0.118595% CI: [-1, 9.1]Cochran-Mantel-Haenszel
Secondary
Advanced Adenoma Detection Rate (aADR)
aADR is defined as the percentage of patients with at least one adenoma ≥ 10 mm, or any adenoma \< 10 mm, which was either of high-grade dysplasia (HGD) or villous or tubulovillous.
Time frame: Day 1
Population: Intent-to-treat
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Colonoscopy (Standard of Care) | Advanced Adenoma Detection Rate (aADR) | 60 Participants |
| CADe Device | Advanced Adenoma Detection Rate (aADR) | 59 Participants |
Comparison: Difference in percentage between both arms is reported. H0: Colonoscopy (Standard of Care) percentage \<= CADe Device percentage (i.e. test of superiority)p-value: 0.819495% CI: [-3.2, 2.6]Cochran-Mantel-Haenszel
Secondary
False Positive Rate (FPR)
FPR is defined as the proportion of colorectal lesions resected and biopsied and subsequently not histologically-confirmed to be clinically relevant colorectal polyps (e.g. a pathology finding of normal mucosa, inflammatory tissue, stool or debris, lymphoid aggregates).
Time frame: Day 1
Population: Intent-to-Treat
| Arm | Measure | Value (LEAST_SQUARES_MEAN) |
|---|---|---|
| Colonoscopy (Standard of Care) | False Positive Rate (FPR) | 0.22 proportion of resected lesions |
| CADe Device | False Positive Rate (FPR) | 0.31 proportion of resected lesions |
p-value: 0.000895% CI: [1.14, 1.69]Exact poisson
Secondary
Flat Adenoma Detection Rate (fADR)
fADR is defined as the percentage of patients with at least one histologically-confirmed non-polypoid adenoma detected.
Time frame: Day 1
Population: Intent-to-Treat
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Colonoscopy (Standard of Care) | Flat Adenoma Detection Rate (fADR) | 19 Participants |
| CADe Device | Flat Adenoma Detection Rate (fADR) | 21 Participants |
Comparison: Difference in percentage between both arms is reported. H0: Colonoscopy (Standard of Care) percentage \<= CADe Device percentage (i.e. test of superiority)p-value: 0.781395% CI: [-1.5, 2]Cochran-Mantel-Haenszel
Secondary
Mean Withdrawal and Inspection Time (MWT)
The withdrawal time is defined as the time measured from the moment the withdrawal phase of the procedure begins (with the scope in the cecum) to the moment the scope is withdrawn from the patient. The Inspection time measurement will exclude washing and resection, and other peri-resection activity not deemed to be colonic inspection. Inspection times for both the control arm and intervention arm will be calculated retrospectively upon review of the video recordings.
Time frame: Day 1
Population: Convenience sample of available videos
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Colonoscopy (Standard of Care) | Mean Withdrawal and Inspection Time (MWT) | 8.97 minutes | Standard Deviation 4.97 |
| CADe Device | Mean Withdrawal and Inspection Time (MWT) | 9.34 minutes | Standard Deviation 4.08 |
p-value: 0.169Permutation test
Secondary
Number of False Alerts Per Procedure
A false alert is defined as a bounding box that persists on the screen (approximately 2-3 seconds per the judgment of the colonoscopist) that is then determined by the colonoscopist not to contain a polyp. The false alert rate is calculated as the number of false alerts per procedure conducted in the intervention arm of the study.
Time frame: Day 1
Population: Intent-to-Treat
| Arm | Measure | Value (NUMBER) |
|---|---|---|
| Colonoscopy (Standard of Care) | Number of False Alerts Per Procedure | 417 Number of false alerts |
Secondary
Polyp Detection Rate (PDR)
Polyp detection rate is defined as the proportion of patients with at least one histologically-confirmed polyp detected.
Time frame: Day 1
Population: Intent-to-treat
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Colonoscopy (Standard of Care) | Polyp Detection Rate (PDR) | 364 Participants |
| CADe Device | Polyp Detection Rate (PDR) | 425 Participants |
Secondary
Polyps Per Colonoscopy (PPC)
PPC is defined as the total number of histologically-confirmed polyps found divided by the total number of colonoscopies performed, per study arm.
Time frame: Day 1
Population: Intent-to-Treat
| Arm | Measure | Value (LEAST_SQUARES_MEAN) |
|---|---|---|
| Colonoscopy (Standard of Care) | Polyps Per Colonoscopy (PPC) | 1.03 Polyps per colonoscopy |
| CADe Device | Polyps Per Colonoscopy (PPC) | 1.32 Polyps per colonoscopy |
Secondary
Proximal Adenoma Detection Rate (pADR)
pADR is defined as the percentage of patients with at least one histologically-confirmed adenoma detected in proximal colon.
Time frame: Day 1
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Colonoscopy (Standard of Care) | Proximal Adenoma Detection Rate (pADR) | 198 Participants |
| CADe Device | Proximal Adenoma Detection Rate (pADR) | 236 Participants |
Comparison: Difference in percentage between both arms is reported. H0: Colonoscopy (Standard of Care) percentage \<= CADe Device percentage (i.e. test of superiority)p-value: 0.057895% CI: [-0.2, 9.4]Cochran-Mantel-Haenszel
Secondary
Serrated Lesions Detection Rate (SLDR)
SLDR is defined as the percentage of patients with at least one histologically confirmed serrated lesion detected.
Time frame: Day 1
Population: Intent-to-Treat
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Colonoscopy (Standard of Care) | Serrated Lesions Detection Rate (SLDR) | 156 Participants |
| CADe Device | Serrated Lesions Detection Rate (SLDR) | 219 Participants |
Comparison: Difference in percentage between both arms is reported. H0: Colonoscopy (Standard of Care) percentage \<= CADe Device percentage (i.e. test of superiority)p-value: 0.000495% CI: [3.7, 12.9]Cochran-Mantel-Haenszel
Secondary
Serrated Lesions Per Colonoscopy (SLPC)
SLPC is defined as the number of histologically confirmed serrated lesions detected, divided by the total number of colonoscopies.
Time frame: Day 1
Population: Intent-to-Treat
| Arm | Measure | Value (LEAST_SQUARES_MEAN) |
|---|---|---|
| Colonoscopy (Standard of Care) | Serrated Lesions Per Colonoscopy (SLPC) | 0.34 serated lesions per colonoscopy |
| CADe Device | Serrated Lesions Per Colonoscopy (SLPC) | 0.46 serated lesions per colonoscopy |
95% CI: [0.05, 0.17]
Secondary
Small Adenoma Detection Rate (sADR)
sADR is defined as the percentage of patients with at least one adenoma 5mm or smaller detected.
Time frame: Day 1
Population: Intent-to-Treat
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Colonoscopy (Standard of Care) | Small Adenoma Detection Rate (sADR) | 201 Participants |
| CADe Device | Small Adenoma Detection Rate (sADR) | 236 Participants |
Comparison: Difference in percentage between both arms is reported. H0: Colonoscopy (Standard of Care) percentage \<= CADe Device percentage (i.e. test of superiority)p-value: 0.075295% CI: [-0.4, 9.2]Cochran-Mantel-Haenszel