Extending the Availability of a Large Multi-Institutional Oncology
Imaging Archive to Remote Reviewers
 
Authors:
Richard E. Hanusik, QARC; Fran Laurie; Thomas J. FitzGerald, MD; Samantha Kelly; Maryann Bishop-Jodoin
 
Background:
The Quality Assurance Review Center (QARC) was established in 1981 to provide radiation therapy (RT) quality assurance (QA) to the Cancer Therapy Cooperative Groups Program.

RT has evolved from 2 dimensional treatment plans, based on external anatomy landmarks, to 3 dimensional, tailored, conformal treatment plans that require CT scans as their basis. Further evolution uses advanced modalities fused with the CT to create the patient’s treatment plan. These advanced imaging modalities are used during treatment to adapt the radiation in real time. Access to the imaging studies used to develop the patient’s treatment plans is required to perform RT QA.

As radiation therapy has evolved to image based, QARC has developed and adapted systems and processes to support the acquisition, management, retrieval, review, and storage of diagnostic imaging.

The QARC system is a comprehensive informatics infrastructure that sustains all the multi-site, multi-disciplinary clinical trial support services that QARC provides. Incorporating quality assurance into trials, through central diagnostic imaging review and interventional RT review, increases compliance and assures uniform study populations. Compliance and uniformity strengthen the validity of study endpoints and improve the standard of care.

 
Evaluation:
Development of the QARC informatics infrastructure began in 1995. Through continued development and evolution of this system, a sophisticated, robust image and RT object management platform is now in use at QARC. A relational database, known as MAX, is the core of the infrastructure. Records on over 47,000 patients enrolled on 359 clinical trials are stored in MAX. Patient records link to associated contacts, benchmarks, clinical trials, correspondence, Digital Imaging and Communications in Medicine data (DICOM) and non-DICOM digital data (referred to as “e-Materials”). MAX features integrated messaging systems which enable users to communicate with each other, the Information Services Division (IS) and external contacts. Patient data is routinely extracted from MAX and sent to the statistical centers of Cooperative Groups and industry partners through established processes. Digital files in MAX include DICOM, DICOM RT and e-Materials. All of the digital files in the Patient Electronic Image Archive (PEI) are linked to the patient’s record in MAX. A user can click on the link from the patient record and the application needed to view the corresponding file is automatically launched and the digital file is displayed. The MAX system is extensible and currently capable of launching 34 different types of files. As patient data is submitted in new formats, the applications needed to view the files are easily incorporated into the system. QARC has undergone a systematic validation process of its information systems to be compliant with FDA 21 CFR Part 11.

QARC uses several image acquisition strategies. The foundational DICOM imaging software system, Dicommunicator, was written by Dr. Keith White of Primary Children’s Medical Center in Salt Lake City, Utah. Dicommunicator is used by institutions worldwide to facilitate DICOM imaging submission to QARC. It has been further developed at QARC and fully integrated into the MAX database. Successful installations of the Dicommunicator software have been performed at remote sites with all major Picture Archiving and Communication System (PACS) vendors. Figure 1 illustrates the MAX Patient Diagnostic Inventory Page and associated imaging studies.
 
Figure 1
Figure 1: MAX Patient Diagnostic Inventory Page
 

The QARC Diagnostic Imaging archive currently contains over 47,000 DICOM studies, including over 12.5 million images. These files have been submitted from over 700 institutions worldwide on 11,292 patients in 139 clinical trials. The imaging modalities represented in the QARC archive include CT, MR, Ultrasound, Mammogram, X-ray, Bone scan, Gallium Scan, PET scan, thallium scan, and MIBG scan. MAX is capable of expanded imaging sets, including spectroscopy and DICOM-compatible pathology objects. These will likely become important data management objectives and tools for cooperative group/industry clinical trials moving forward. Figure 2 demonstrates the growth of the QARC PEI. The growth is represented in two curves, one for DICOM studies, and the other for e-Materials. The e-Material curve in Figure 2 includes radiation therapy treatment plan files submitted in either DICOM RT or RTOG data formats.
 

Figure 2
Figure 2: Growth of the QARC PEI
 
Figure 3
Figure 3: QARC Network Access Diagram
 
Discussion:
A major strength of the QARC program, and one of the main reasons for its growth, is the diversity of successful methods for acquiring digital data. QARC has developed processes to support diversity in image transfer. Digital data required for clinical trials are submitted via CD, Dicommunicator, and Secure File Transfer Protocol (sFTP). These are readily imported into the QARC PEI, linked to the patient records in MAX, and available for review.

QARC Supports both radiotherapy and imaging reviews. Radiotherapy reviews evaluate the treatment plan to assure compliance with clinical trial guidelines. Real-time imaging reviews confirm response and determine the next treatment path. In 2008 QARC on-site radiologists and radiation oncologists performed 3051 real-time imaging and radiotherapy reviews. Web-based conferences are routinely held between reviewers at QARC and remote sites to collaborate and resolve discordance between central and local review. Cases may undergo multiple reviews. Studies have shown that if modifications identified by interventional review were required and made, the treatment became compliant with clinical trial requirements.

Current treatment regimens on cooperative group clinical trials require real-time response assessments with central review confirmation. Many of these studies are large, requiring participation from more than 250 treating institutions. Historically, central reviews have been performed on-site at QARC. More recently, we have developed the technologies to extend the availability of this large multi-institutional imaging archive to remote reviewers.
 
Conclusion:
Development to provide the functionality and ease of use of the QARC system to remote reviewers began in 2008. Using terminal server and remote desktop technologies, remote reviewers have the ability to access patient records and associated imaging via a secure connection using an internet browser from anywhere in the world, as illustrated in Figure 3. This has expanded the service provided by QARC. To date, QARC has facilitated over 1,000 diagnostic imaging reviews using the remote reviewer functionality. When the remote reviewer connects to QARC and their credentials are recognized by MAX, they are presented with the specific work list of pending reviews. From the work list, the remote reviewer has access to a streamlined version of the patient record with the links in place to launch diagnostic imaging or e-Materials. The data fields in the patient record are locked and cannot be edited by the remote reviewer. DICOM imaging is displayed using a two-study viewer, so that images can be evaluated side by side. Images can be annotated and saved, and evaluations and comments can be recorded in the patient record. Audit trails for the remote review process are in place.

The QARC archive may be one of the largest multi-site, multidisciplinary oncology archives. QARC data management systems have met the needs of complex and advanced technology trials. As clinical trials incorporate advanced technologies, quality assurance and real time review become more critical. Establishing an integrated database for distributed remote review of imaging, radiation therapy, and pathology objects offers a unique opportunity to make a vast array of patient data available to the clinical research community.

This work was supported by NIH/NCI U10 Grant CA29511

 
References:
1. Laprise N, Hanusik R, FitzGerald TJ, Rosen N, White K. Developing a Multi-Institutional PACS Archive and Designing Processes to Manage the Shift from a Film to a Digital-Based Archive. J Digit Imaging. 2009;22:15-24.

2. FitzGerald TJ, Urie M, Ulin K, et al. Processes for Quality Improvements in Radiation Oncology Clinical Trials. Int J Radiat Oncol Biol Phys. 2008;71(S1):S76-9.

3. Peters L, O’Sullivan B, Giralt J, et al. Trans Tasman Radiation Oncology Group. Critical Impact of Radiotherapy Protocol Compliance and Plan Quality in the Treatment of Advanced HNSCC: Results From TROG 02.02, Barcelona, Spain. February 2009.