Plain English Summary
Background and study aims
68Ga-PSMA-11 PET/CT is a new imaging technique developed to use for investigating how advanced a cancer is and where it is located. It can be used, for example, to monitor the response of a cancer to radiotherapy. The overall value of a imaging technique such as this depends on the degree of reader agreement – that is, if a number of people agree on what the image is showing. Knowing how different people will interpret the same image (interobserver variability) and how many people will agree that it shows the same thing (reproducibility) is essential for interpreting the results of studies and for designing new studies. The aim here is to determine interobserver agreement for 68Ga-PSMA-11 PET/CT interpretations for prostate cancer and to compare the findings among highly experienced and less experienced readers.
Who can participate?
Nuclear medicine physicians or radiologists with PET/CT experience.
What does the study involve?
Participants are given 68Ga-PSMA-11 PET/CT images from anonymous patients with prostate cancer and, upon a visual inspection, are asked to report on their findings using a standard template. They are asked to interpret what they see from a number of images of lymph nodes affected by localized prostate cancer and divided into different regions or sections. The participants are asked to judge whether each image they see shows prostate cancer. The results are compared to that of histopathology and follow-up imaging taken for all the patients more than six months before the 68Ga-PSMA-11 PET/CT imaging. Interobserver agreement between participants is assessed.
What are the possible benefits and risks of participating?
Not provided at time of registration
Where is the study run from?
Department of Nuclear Medicine, Ludwig-Maximilians-University Munich (Germany)
When is the study starting and how long is it expected to run for?
October 2016 to March 2017
Who is funding the study?
Department of Nuclear Medicine, Ludwig-Maximilians-University Munich (Germany)
Who is the main contact?
Dr Wolfgang Fendler
Trial website
Additional identifiers
EudraCT number
ClinicalTrials.gov number
Protocol/serial number
13102016.1.0
Study information
Scientific title
68Ga-PSMA-11 PET/CT interobserver agreement for prostate cancer assessments: an international multicenter prospective study
Acronym
PSMAGREE
Study hypothesis
The overall value of an imaging method is associated with the degree of reader agreement. Knowledge of interobserver variability and reproducibility is therefore essential for interpreting study results and design of future trials. The aim of this study is to determine interobserver agreement for interpretations of 68Ga-PSMA-11 PET/CT and to compare findings among readers with low and high levels of experience.
Ethics approval
Ethics Committee of the Ludwig-Maximilians-University Munich, 10/06/2016, ref: 594-16UE
Study design
Observational
Primary study design
Observational
Secondary study design
Cross sectional study
Trial setting
Hospitals
Trial type
Diagnostic
Patient information sheet
Condition
Prostate cancer
Intervention
Anonymized PET/CT images of 50 patients included in the study (one scan per patient) will be electronically submitted to more than ten nuclear medicine physicians/radiologists from Europe, Asia, Australia or North America. Data include standard DICOM files of CT and attenuation-corrected PET images.
All reader report a visual interpretation of the 68Ga-PSMA-11 PET/CT image datasets using a standard template. Visual interpretation will be performed for locoragional lymph nodes - divided into pelvic regions, extra-pelvic lymph nodes, bone and visceral organs. Each region will be judged positive or negative for prostate cancer involvement.
For binary data, agreement among each observer and the reference reader will be evaluated using Cohen's κ. Overall agreement using pooled observer data will be evaluated using generalized estimation equations. For non-binary data, agreement among all observers was evaluated by intraclass correlation coefficient (ICC) using two-way mixed model for absolute agreement (single measures). Ninety-five percent confidence intervals (CIs) are reported for κ and ICC values. Interpretation of κ and ICC will be based on a classification provided by Landis and Koch: 0.0, poor; 0.0–0.20, slight; 0.21–0.40, fair; 0.41–0.60, moderate; 0.61–0.80, substantial; 0.81–1.00, almost-perfect reproducibility. Discrepancies in quantitative ratings among observers will be expressed as mean difference (Δ) ± standard deviation (SD). Sensitivity, specificity, positive predicive value, and negative predicive value will be calculated for each observer
Histopathology and follow-up (PSA and imaging) obtained for all patients more than six months before the index test serves as reference standard. Lymph nodes regions, organs and bone will be judged positive or negative for prostate cancer involvement based on the results from histopathology and follow-up imaging.
Intervention type
Device
Phase
Drug names
Primary outcome measures
Interreader agreement, i.e. the variance between imaging findings of different readers, calculated by Cohen's κ.
Secondary outcome measures
1. Accuracy for each observer compared to the reference standard, determined by sensitivity, specificity, positive predicive value, and negative predictive value
2. Comparison of accuracy between readers of low and high levels of experience, by difference in pooled sensitivity, specificity, positive predictive value, and negative predictive value
Overall trial start date
13/10/2016
Overall trial end date
01/03/2017
Reason abandoned
Eligibility
Participant inclusion criteria
Nuclear medicine physician or radiologist with PET/CT experience.
Participant type
Health professional
Age group
Adult
Gender
Both
Target number of participants
16
Participant exclusion criteria
Physicians with prior knowledge of the included patient datasets.
Recruitment start date
01/11/2016
Recruitment end date
13/12/2016
Locations
Countries of recruitment
Australia, Austria, Denmark, France, Germany, Japan, United States of America
Trial participating centre
Department of Nuclear Medicine, Ludwig-Maximilians-University Munich
Klinik und Poliklinik für Nuklearmedizin
Klinikum der Universität München
Marchioninistraße 15
81377 München
Munich
81377
Germany
Trial participating centre
University of California Los Angeles
Ahmanson Biological Imaging Clinic
10833 LeConte Ave.
Los Angeles
CA 90095-6948
United States of America
Trial participating centre
University Hospital Essen
Department of Nuclear Medicine
Essen
-
Germany
Trial participating centre
Hokkaido University Graduate School of Medicine
Department of Nuclear Medicine
Kita 15, Nishi 7, Kita-ku
Sapporo
060-8638
Japan
Trial participating centre
Technical University of Munich (TUM)
Department of Nuclear Medicine
Munich
-
Germany
Trial participating centre
University of Ulm
Department of Nuclear Medicine
Ulm
-
Germany
Trial participating centre
Bichat University Hospital
Department of Nuclear Medicine
Inserm 698
University Paris 7
46 rue Henri Huchard
Paris
75018
France
Trial participating centre
University of Rostock
Department of Nuclear Medicine
Rostock
18051
Germany
Trial participating centre
St Vincent's Public Hospital Sydney
Department of Nuclear Medicine
Sydney
-
Australia
Trial participating centre
Aalborg University Hospital
Department of Nuclear Medicine
Hobrovej 18-22
Aalborg
9000
Denmark
Trial participating centre
University of California San Francisco
Department of Nuclear Medicine
05 Parnassus Ave
San Francisco
CA 94143
United States of America
Trial participating centre
General Hospital of Vienna (AKH)
Department of Nuclear Medicine
Vienna
-
Austria
Trial participating centre
University Hospital of Würzburg
Department of Nuclear Medicine
Würzburg
-
Germany
Funders
Funder type
University/education
Funder name
Department of Nuclear Medicine, Ludwig-Maximilians-University Munich
Alternative name(s)
Funding Body Type
Funding Body Subtype
Location
Results and Publications
Publication and dissemination plan
Results (Cohen's κ, intraclass correlation coefficient, sensitivity, specificity, positive predicitive value, negative predictive value, accuracy) will be published within three to six months of completion of the trial.
Intention to publish date
01/03/2017
Participant level data
Not expected to be available
Results - basic reporting
Publication summary