The principal role of the American College of Radiology Imaging Network (ACRIN) is to conduct multicenter clinical trials of diagnostic imaging and image-guided treatment technologies as they are used in cancer. Soon, however, ACRIN will assume a new responsibility. ACRIN is collaborating with the Academy of Molecular Imaging (AMI) to develop a registry for PET examinations that will allow PET centers to bill Medicare for cancer-related studies that currently are not reimbursable. The idea to develop the registry is the result of an early 2005 announcement by the Centers for Medicare and Medicaid Services (CMS) that the agency wished to deal with PET reimbursement through its new “coverage with evidence development” program. Under this program, CMS will pay for applications of PET to cancer diagnosis, staging, and treatment monitoring that are not among those indications currently approved, so long as data is submitted to a CMS-approved registry or clinical trial. The goal of the program is to develop sufficient data to help CMS make more definitive PET coverage decisions.

Dubbed the National Oncologic PET Registry (NOPR), the registry will address important problems currently being experienced by cancer patients, referring physicians, and radiologists. Because Medicare PET coverage is spotty, patients who might benefit from PET scans are either not having them or paying for these expensive examinations out of pocket. Referring physicians are caught in the middle of having to balance their enthusiasm for how PET might improve their patients’ management with their patients’ financial concerns. Many PET centers have excess and unused capacity that could be addressed by more expansive coverage for cancer indications. Given these concerns, and that the traditional path to Medicare reimbursement through generalizable, multicenter clinical research might take many years, the NOPR is supported by the American College of Radiology, the American Society of Clinical Oncology (ASCO), and the Society of Nuclear Medicine (SNM).

PET centers will register to participate in NOPR via a Web site, www.CancerPETregistry.org, and pay a small registration fee. Subsequently, upon referral of eligible patients to the PET facility, the referring physician must complete survey forms prior to PET scanning and after receiving the PET results. The survey forms question the referring clinician about aspects of the patient, their disease process, specifically why the study is being requested, and the intended management plan. Differences between therapeutic expectations prior to the scan and following PET form the underpinnings for the analysis, which is based on several previously published, single institutional studies. Timely completion of the forms will trigger Medicare reimbursement. PET facilities will pay a small fee for each patient that goes to support the activities of the registry. All data will flow over the Web, using software designed for the purpose by the ACR, to ACRIN headquarters in Philadelphia, where it will be permanently archived. The initial analysis will be performed by the ACRIN-AMI working group members, who will advise CMS of the results periodically as data accrues. Ultimately, the data will be available to CMS and others who wish to perform secondary analyses.

The principal assessment of the data will focus on demonstrating the magnitude of the impact of employing PET on referring physician decision-making for individual disease site/clinical indication pairs. For example, employing PET for treatment monitoring of patients with non-small cell lung cancer (NSCLC) would be one evaluation, while PET for diagnosis or initial staging of sarcoma would be another. It is possible, even likely, that CMS eventually will use the resultant data to make separate coverage decisions for each disease site/indication pair. The design allows for analysis of pairs of 30 cancer types and four principal indications (diagnosis, initial staging, suspected recurrence or restaging, and treatment monitoring); 23 disease site/indication pairs already have full or partial CMS approval for reimbursement. Among the remaining pairs, a number of disease site/indication pairs have been listed as “high priority,” including: therapeutic monitoring for lymphoma, NSCLC, metastatic colorectal cancer, head and neck cancer, and esophageal cancer; diagnosis of pancreatic cancer and cancer of unknown origin; initial staging of small cell lung, pancreatic, and ovarian cancers and malignant melanoma; and restaging or suspected recurrence for ovarian and cervical cancers, brain tumors, and multiple myeloma.

Most of the hurdles to implementing the registry have been surmounted. Since the data is being collected only to provide advice on Medicare coverage, it is not considered research, and the expectation of CMS is that institutions will not have to obtain Institutional Review Board approval. The Office of Management and Budget is studying the financial impact. There is hope that the NOPR will be activated by early 2006.

Bruce J. Hillman, MD, is the Theodore E. Keats Professor of Radiology, University of Virginia, Charlottesville, and chair, American College of Radiology Imaging Network (ACRIN).