PET: New indications give PET a lift after struggle over regulatory and reimbursement issues
The nuclear medicine community has developed innovative ways of improving access to this technology, including mobile systems
By Peter S. Conti, M.D., Ph.D., and Jennifer Keppler, CNMT, MBA
PET imaging using the radiotracer fluorine-18 fluorodeoxyglucose provides a unique means for metabolic characterization of neurological, cardiac, and oncologic disease. In particular, the FDG-based whole-body PET technique developed in recent years has surpassed most expectations of its utility in clinical oncology. The large spectrum of neoplasms that can be studied makes this an essential imaging tool in the diagnosis and management of many patients with cancer.
PET images functional processes, such as blood flow or metabolism, in vivo. Most PET scans rely on use of the glucose analog F-18 FDG, which, following intravenous injection, distributes in the body in proportion to tissues’ glucose consumption. Most cancers, and certain normal organs such as the brain, exhibit markedly increased rates of glucose utilization, resulting in prominent accumulation of the radiolabeled drug in those tissues. In addition to dedicated imaging capabilities involving the heart and brain, whole-body PET provides a means of evaluating disease processes such as cancer.
The number of PET scanners available for clinical use in the U.S. has increased over the last decade, but not to the extent the technique’s clinical utility would indicate. There are about 200 PET systems in clinical use in the U.S. Dedicated state-of-the-art PET systems are still relatively expensive, ranging in price from $800,000 to $2.5 million. Smaller medical communities have been hesitant to make such a large investment; the expense of the equipment and infrastructure necessary for an on-site medical cyclotron to produce the radioactive materials necessary for PET radiopharmaceuticals has also discouraged expansion of the modality.
In response to these issues, the nuclear medicine community has developed a number of innovative approaches to improve access to this technology. These include development of lower cost imaging systems that allow the use of scanners for both conventional single-photon emission computed tomography (SPECT) and PET imaging. In addition, creation of radioisotope distribution centers throughout the U.S. provides hospitals with access to radiopharmaceuticals such as FDG without the need for capital investment in cyclotrons. The use of mobile systems also brings state-of-the-art PET scanners into smaller communities without the need for large capital investment in imaging equipment (see sidebar story on page 44).
Clinical Applications
The role of PET and F-18 FDG in imaging certain neurological disorders, including epilepsy, stroke, dementia (particularly Alzheimer’s disease), and movement disorders, along with cardiac perfusion and viability, has been recognized for some time. Acceptance of PET’s role in cancer diagnosis, staging, and treatment monitoring has grown in the last few years. Data now exist to support the clinical use of PET and F-18 FDG in cancers of the brain, head and neck, lung, breast, esophagus, colon, pancreas, ovary, musculoskeletal system, skin, and lymphatic system, with newer applications rapidly emerging in the literature. Most oncology patients who undergo PET scanning have either lung or colorectal cancer.
The latest applications approved for reimbursement by Medicare are lung cancer, colorectal cancer, lymphoma, and melanoma.
Lung cancer. Patients with suspected lung cancers are traditionally evaluated by planar chest x-rays, CT scanning, and occasionally by MRI. While certain parameters on radiographic studies indicate a likelihood that a solitary pulmonary nodule is malignant, a substantial portion of such lesions are radiographically indeterminate. Imaging characteristics that increase the probability of malignancy are size, wall thickness of cavitary lesions, absence of calcification, irregular margins, and growth on follow-up studies. Although such findings are helpful, there is a low specificity associated with standard radiographic criteria, resulting in a high frequency of invasive procedures to corroborate imaging results through pathological assessment.
Since benign disease is more frequently found in solitary pulmonary nodules biopsied in the general population, it could be argued that many patients are exposed to unnecessary morbidity. Characterization of pulmonary lung nodules with PET has been extensively evaluated with an overall sensitivity, specificity, and accuracy of 95%, 88%, and 94%, respectively, for detection of a malignant primary tumor. The cost reduction due to avoidance of invasive procedures has been estimated at $30 million to $236 million per year in the U.S. for this application alone. Many practitioners use PET routinely in the workup of such lesions rather than proceeding directly to biopsy.
Likewise, PET has contributed to the staging of patients with suspected or known lung cancer. Although conventional imaging techniques find 20% to 30% of lung masses to be resectable by current criteria, 5% to 7% of the patients are found at surgery to have unresectable disease. Furthermore, 14% of all patients who have surgery die within a year of what was thought to be a curative surgical procedure. This is likely related to the fact that the morphologic imaging techniques have well-known limitations in identifying metastatic lesions, including those in mediastinal lymph nodes. Instead of surgery, patients with lung cancer who are more accurately and noninvasively staged by PET experience a lower incidence of mortality and morbidity, have lower medical costs since surgery is eliminated and potentially have an improved quality of life. In addition, chemo- or radiotherapy, rather than surgery, may be indicated if mediastinal disease is present or if the risk of mortality is high, as in the elderly or patients with obstructive airway disease.
The use of PET for staging lung cancer in the chest has resulted in altered staging in about 40% of patients studied, many of whom become inoperable based on the presence of previously unsuspected extensive disease (Figure 1). More accurate staging also means that detection of disease outside of the chest is possible with this whole-body evaluation. Recent work has also demonstrated a role for PET in detection of recurrent disease and in radiation treatment planning.
Colorectal cancer. The most significant diagnostic imaging challenge in colorectal carcinoma occurs in the post-therapy patient suspected of having recurrence. About 20% to 50% of patients who undergo potentially curative surgery develop local recurrence within the following two years, with a large portion of deaths occurring secondary to these loco-regional recurrences. The low specificity of CT scanning and of the serum markers such as CEA in these patients frequently necessitates biopsy of suspected recurrences. A negative biopsy result, however, cannot exclude recurrence.
Results with PET demonstrate about 95% sensitivity, 98% specificity, and 96% accuracy in patients with suspected local recurrence. PET has led to substantial changes in patient management, and its advantage in assessing patients with suspected recurrence appears clear-cut when compared with other diagnostic approaches, including immunoscintigraphy. Data are also available to demonstrate the role of PET in whole-body evaluation of metastatic disease in patients being considered for loco-regional resection of recurrent disease. Identification of distant abdominal or thoracic disease is clearly important when considering surgical versus medical intervention (Figure 2).
Lymphoma. In the past, CT scanning and gallium scintigraphy led to marked improvement in the ability to accurately stage and follow patients with lymphoma. Anatomically based imaging modalities have considerable shortcomings, however, particularly their lack of specificity when findings are abnormal. The limitations of current anatomic imaging studies are most evident in the post-therapy patient. An enlarged lymph node initially containing tumor, for example, may remain enlarged permanently after effective eradication of tumor by therapy. Gallium-67 has been used successfully to stage patients with high-grade lymphoma, but this technique has limitations in resolving small or low-grade lesions and in evaluating the abdomen, as in detection of liver disease.
Data in the literature demonstrate that PET appears to be more accurate than CT in pre- and post-therapy staging of lymphoma. In addition, it has been shown that FDG may be more useful than gallium scintigraphy in evaluating the various grades of malignancy, and that patient management can be altered with the addition of PET in 5% to 20% of patients with lymphoma (Figure 3). The diagnostic role for PET has also been established for brain lesions in patients with AIDS, in whom it can be used to provide accurate differentiation of primary CNS lymphoma from non-neoplastic lesions such as toxoplasmosis. Finally, there is emerging evidence to suggest that PET can be used to assess treatment response. Traditional means of assessment require substantial delays to ascertain clinical response or alterations in morphological appearance of tumor masses. Alterations in FDG accumulation as a result of therapy could be used at earlier time points in determining therapeutic effect.
Melanoma. The incidence of malignant melanoma is rapidly increasing, with early diagnosis dependent on visual surveillance of at-risk individuals. This is often followed by surgical biopsy of suspicious lesions and subsequently by wide surgical excision of identified melanomas.
The high mortality of melanoma is due to its propensity for early nodal metastasis and not infrequently to hematogenous visceral spread. A noninvasive technique capable of identifying patients with nodal metastases could provide a more rational basis for recommending nodal dissection and sampling, procedures that are often costly and not infrequently morbid.
Due to its increased sensitivity for lesion detection, PET has shown an obvious advantage over other imaging methods such as CT and gallium scintigraphy. Several studies have suggested that PET as a single study can stage melanoma patients and screen for recurrence more accurately than conventional methods (Figure 4). In the brain, however, as with other forms of metastatic cancers, the higher anatomic resolution of MRI and the normally high uptake of FDG in the gray matter appear to make MRI more useful than PET in screening for small metastases in neurologically asymptomatic patients. Patient management is also changed after PET scans in many patients with melanoma: Surgery is frequently canceled because of the presence of extensive unresectable disease, or redirected to additional areas to attempt a surgical cure.
Regulatory Challenges
The advancement of PET has become possible in large measure because of the resolution of many regulatory challenges. Food and Drug Administration regulation is critical in medicine because providers and purchasers of services do not equally understand the benefits of various choices. FDA approval provides assurance that a technology or drug is safe and effective for its proposed use. In many ways, such regulation provides the structure and stability necessary for an industry to flourish.
Despite the need for them, regulators are not always able to provide systems that work perfectly. Although practitioners accustomed to PET were convinced of its value, proliferation could not occur until this and other regulatory challenges could be addressed. For many years, the field of PET was mired in a regulatory process that did not work. The FDA Modernization and Accountability Act of 1997 (FDAMA) brought about changes that have refined and improved the regulations for PET.
FDAMA contains specific language requiring the FDA to adopt appropriate procedures for approval of new drug applications (NDAs) and abbreviated NDAs (ANDAs) for PET radiopharmaceuticals, as well as appropriate current good manufacturing practices (CGMPs) for the production of these PET compounds. In addition, the FDA is required to take “due account of any relevant differences” between commercial PET centers and not-for-profit PET facilities to reduce the burden of coming into compliance for non-commercial PET production facilities. The new requirements are to be determined by the FDA, in consultation with industry, patients, and the user community.
A PET radiopharmaceutical committee (PET RC), formed under the initiative of the Institute for Clinical PET (ICP) with representatives of the Society of Nuclear Medicine and other organizations, has been assisting the FDA in developing the regulations mandated by FDAMA. Efforts have focused on completing safety and efficacy evaluation of PET radiopharmaceuticals in clinical use, and on developing chemistry guidelines for NDAs and manufacturing practices. The new regulatory approach was expected to be published in the form of a proposed rule in the Federal Register in December 1999.
Guidance documents will be published in the future that will assist the community in coming into compliance over the following, mandated two-year time period. The proposed rule will identify the new PET compounds for which safety and efficacy has been established, describe their labeled use, and define a process by which PET production sites can manufacture these drugs. A summary of the progress to date on the proposed rule follows.
Safety and efficacy of PET compounds. The first step in bringing PET into regulatory compliance is to determine which PET drugs can be approved for use. Because of the lack of a sponsor for a single drug approval application, the FDA determined that it would conduct the safety and efficacy evaluations of PET compounds in common clinical use around the country. It had previously approved two positron-emitting drugs: rubidium-82 for perfusion imaging of the heart, and F-18 FDG for evaluation of epilepsy. The PET RC targeted FDG for additional indications, as well as nitrogen-13 ammonia (NH3), oxygen-15 water, F-18 fluoride, and F-18 fluoro-dopa for evaluation.
FDA evaluation of these compounds was to culminate in the publishing of the Federal Register announcement noted above, detailing the findings of the analysis and requesting applications for production of these drugs. The evaluation would be completed on the basis of published literature, a so-called “paper” NDA.
The FDA completed preliminary analyses of FDG, NH3, and water, in consultation with the PET community and presented the findings to the FDA’s Medical Imaging Drug Advisory Committee in June 1999. The advisory committee found that FDG is safe and effective in PET imaging for assessment of abnormal glucose metabolism to assist in the evaluation of malignancy in patients with known or suspected abnormalities found by other testing modalities, or in patients with existing diagnoses of cancer.
The committee also found FDG safe and effective with PET imaging in patients with coronary artery disease and left ventricular dysfunction, when used together with myocardial perfusion imaging to examine myocardial glucose metabolism and to identify myocardium with reversible loss of systolic function. In addition, the committee found NH3 safe and effective in PET imaging of the myocardium under rest or pharmacological stress conditions to evaluate myocardial perfusion in patients with suspected or existing coronary artery disease.
The advisory committee found it necessary to review additional published literature on the validation of water as a brain perfusion agent and the FDA agreed to gather and analyze this literature. In conjunction with the PET community, an analysis of the literature in support of the clinical use of FDG in dementia and F-dopa for movement disorders will be undertaken. Once these are completed, the Medical Imaging Drug Advisory Committee will be reconvened this spring to evaluate these PET drugs. This same process, managed in the future by the PET community, will remain as one mechanism for obtaining FDA approval of new PET radiopharmaceuticals.
Chemistry and manufacturing regulations. Establishing the safety and efficacy of PET compounds is just one step in bringing PET into regulatory compliance. The FDA must also assure that the sites manufacturing these drugs produce high-quality compounds on a continual basis. Typically, this is done through the registration of manufacturing sites, filing of applications to manufacture the drugs and enforcement of standard manufacturing practices.
At present, the FDA asserts that all PET production sites will need to register as drug establishments and list the drugs in clinical use (for which there is an exchange of value for their use). This would require that sites file NDAs or ANDAs for the production of the compounds, follow current good manufacturing practices that are being developed for PET, and be open to FDA inspection. While many in the community reference regulatory guidelines stating that academic institutions and hospitals do not need to register as drug manufacturing institutions, the FDA contends that FDAMA mandates compliance with FDA oversight.
At a public meeting in February 1999, the agency presented a summary of the existing regulatory process for manufacturing drugs and some preliminary information as to how these may be implemented for PET. The FDA revealed plans to develop templates for the community to use in applying for NDAs/ANDAs. Since the safety and efficacy of PET drugs was to be established through the Federal Register announcement described above, NDAs and ANDAs will consist solely of these templates for chemistry, manufacturing, and controls (CMC). Draft templates for production of FDG, NH3, and the fluoride ion have been developed in consultation with the PET community and are available on the FDA’s Web site.
A site applying to produce any of these drugs will reference the Federal Register announcement, fill out a CMC application, and answer a few additional questions. The process has been simplified substantially over the prior regulatory framework.
The remaining issue for regulation of PET is the development of appropriate manufacturing practices that can be used as standards for facilities producing these compounds. The FDA distributed a draft of potential CGMPs at a public meeting in September 1999.
Reimbursement
While the new regulatory processes were being developed for PET, the Health Care Financing Administration reconsidered its national coverage policy on PET. At the time the first compound, Rb-82, was approved, HCFA issued a national policy statement restricting coverage for Medicare patients to Rb-82 and specifically stating that other indications for PET were considered experimental. Thus, unlike most new technologies, all coverage decisions for PET will be controlled by HCFA. Local carrier discretion would not be allowed as a venue to expand Medicare coverage of PET.
At the time FDAMA was passed, HCFA revised its policy, requiring that Medicare beneficiaries have access to PET scans for characterization of solitary pulmonary nodules and initial staging of lung cancer, beginning Jan. 1, 1998. A fast-track review of other indications for PET—evaluation of brain tumors, myocardial viability, colorectal cancer, head and neck cancer, melanoma, breast cancer, Hodgkin’s lymphoma, and ovarian cancer—would be initiated to expand this coverage.
To facilitate the reimbursement/approval process, HCFA held a “town hall” meeting in January 1999 to evaluate several potential new PET indications. Because HCFA technology evaluation was in a state of flux at the time of the meeting, no official recommendation could be made; however, HCFA staffers would be able to make coverage decisions based on the information presented.
The meeting focused on discussions of the clinical data supporting the use of PET imaging in five potential indications: colorectal cancer, melanoma, head and neck cancer, lymphoma, and brain tumors. Members of the community, under the direction of the ICP, organized a complement of presentations on these topics. PET practitioners, surgeons, and oncologists presented data on PET’s usefulness in the care and management of their patients. Supplementing these presentations were the testimonies of 10 patients whose lives had been affected by PET.
Discussions with HCFA centered on the design of the clinical trials that were used in many of the PET studies. The question was whether randomized controlled trials and outcome studies showing increased quality and survival were still necessary for PET to be approved by HCFA. Proponents of PET argued that randomization is not needed because patients serve as their own control and that outcome for diagnostics is related to results of the test, not improvements in life expectancy.
In March 1999, HCFA announced that Medicare’s coverage policy would expand, effective July 1, to include three new indications for whole-body PET scans using F-18 FDG:
evaluation of recurrent colorectal cancer in patients with rising levels of carcinoembryonic antigen (CEA);
staging and characterization of lymphoma, (both Hodgkin’s and non-Hodgkin’s lymphoma, when done as an alternative to a gallium scan); and
detection of recurrent or metastatic melanoma prior to surgery.
Subsequent to the announcement of coverage for the three additional indications, HCFA published a notice in the Federal Register on April 27, 1999, describing its process for making national coverage decisions. This new process, which became effective in June 1999, will be used to expand Medicare coverage of PET.
Future Outlook
In addition to its more established roles in neurological and cardiac disorders, the PET FDG technique has now been applied to a large spectrum of neoplasms. Research and clinical experience have demonstrated its role as a key tool in diagnosis and management for many patients with cancer. In most circumstances, diagnostic sensitivity, specificity, and accuracy are 85% to 90% or better whether addressing primary lesions, metastases, or recurrences. By comparison, anatomically based modalities generally fail to display such impressive percentages.
These factors, together with the imminent resolution of the regulatory and reimbursement hurdles, will provide patients and their physicians the opportunity to access the power of this technology into the next century.
Dr. Conti is medical director of the PET Imaging Science Center at the University of Southern California in Los Angeles, and president of the Institute for Clinical PET. Ms. Keppler is executive director of the ICP.
Bibliography
Conti PS, Keppler J, Halls JM. Positron emission tomography: a financial and operational analysis. AJR 1994;162:1279-1286.
Conti PS. Introduction to imaging brain tumor metabolism with positron emission tomography. Cancer Invest 1995;13:244-259.
Conti PS, Lilien DL, Grafton ST, et al. PET and [18F]-FDG: a clinical update. Nucl Med Biol 1996;23:717-735 (and references therein).
Keppler JS, Thornberg CF, Conti PS. Regulation of positron emission tomography: a case study. AJR 1998;171:1187-1192.
Sidebar
Mobile PET follows tracks of MRI
Small hospitals can now offer PET services
By Johannes Czernin, M.D.
The revolution occurring in nuclear medicine is one of the most exciting aspects of the changing world of diagnostic imaging. The development of whole-body positron emission tomography (PET) together with the imaging agent fluorine-18 deoxyglucose (FDG) allows the accurate detection and staging of cancer.1,2 The technology has resulted in a convincing body of knowledge that can be applied to benefit patients with cancer, neurological disorders, and heart disease.
The number of PET scanners in use is rapidly increasing. Nuclear medicine technology has adapted to the increased demand by designing coincidence imaging systems that allow the use of scanners for both conventional SPECT and PET imaging. Radioisotope distribution centers have been established throughout the country to provide hospitals with the opportunity to perform metabolic imaging. These centers allow for distribution of radiotracers such as FDG, thus permitting smaller medical communities to benefit from PET technology.
Dedicated state-of-the-art PET systems range in price from $800,000 to $2.5 million. Smaller institutions are hesitant to commit to such substantial investments. This has led to mobile systems, which have successfully brought CT and MRI to these communities. The concept is fairly simple. Radioisotope distribution centers, that is, cyclotrons that produce FDG, need to be available within a reasonable distance of the user. Because the physical half-life of FDG is about two hours, travel times longer than that are undesirable.
The dedicated PET unit is integrated into a specifically designed trailer or coach and is operated by PET technologists who coordinate and implement the services. The mobile PET unit requires a special pad and shore power for proper operation of the tomograph. Many institutions fulfill this requirement because they are routinely utilizing the services of mobile CT, MRI, or lithotripsy units.
The user orders the radioisotope (FDG) from a regional supplier. The local technologist in the nuclear medicine department injects the tracer. After an uptake period of about 30 minutes, the patient is transported to the mobile unit, where images are acquired. The imaging procedure takes about 30 minutes for FDG studies of the brain and one hour for whole-body tumor surveys. While one patient is being imaged, the next patient can be injected and imaging can start immediately upon completion of the first patient. In this way, eight to 10 patients can be imaged during an eight-hour working day.
Typically, mobile companies bill the medical institution and the institution bills third-party payers. Because of the costs for operating the mobile unit, users usually perform a minimum of four patient studies a day. Typical fees for mobile PET services range from $1200 to $1600 per patient study, including the costs for the radioisotope. Medicare reimbursement for the approved applications ranges from $1900 to $2500. Thus, even smaller institutions can provide state-of-the-art PET services at a minimal financial risk.
PET imaging has emerged as an important clinical tool for managing patients with oncological, neurological, and cardiological diseases. Large hospitals have established PET centers to improve the medical care of those patients. Now, mobile PET providers allow small hospitals and physicians’ groups to offer PET services to their community.
Johannes Czernin, M.D. is the director of pharmacology/nuclear medicine at University of California, Los Angeles.
