Dr. Tan, Mr. Turner, and Dr. Forde are with the Molecular Imaging Institute, Radiologic Associates of NW Indiana, Inc., Crown Point, IN. They are also affiliated with the Department of Radiology and Dr. Klein and Dr. Tabib are with the Department of Oncology at Porter Memorial Hospital, Valparaiso, IN.

Since this article was written, Dr. Tan is now with the Department of Radiology, St. Rita's Medical Center, Lima, OH.

Since our previous review of clinical applications of positron emission tomography (PET) in this journal in 1995, there have been many developments in PET technology, clinical experience, and reimbursement policies.

There are three major areas in which PET scanning has proven useful: oncology, cardiology, and neurology. ^1-3 The vast majority of PET scanning performed in the United States is in oncology. There are two reasons for this trend: clinical impact and reimbursement policy. In recent years, the results of large clinic PET studies involving tumor detection and staging have provided solid evidence of the cost-effectiveness of this modality. This has a major impact on clinical decision-making, with the effect of impacting patients' treatment plans. ^4-48 This scenario, in turn, enhanced reimbursement policy dev-elopment on indications for oncologic PET imaging.

In contrast, neurologic and cardiologic PET imaging has lagged behind more than had been expected. For instance, one of the major applications of neurologic PET imaging is the accurate diagnosis of Alzheimer's disease. However, since no effective preventive nor therapeutic measures are available for this disease, the use of PET for diagnosis cannot have a clinical impact on patient outcome.

Cardiologic PET imaging is used almost exclusively for myocardial viability assessment. This requires PET myocardial perfusion with N-13 ammonia combined with PET metabolism with F-18 fluorodeoxyglucose (FDG) to detect mismatched perfusion-metabolism (diminished myocardial perfusion with preserved metabolism). N-13 ammonia has a half-life of 10 minutes and requires an expensive onsite cyclotron (at a cost of $2 to $2.5 million) to produce it. Because of its short half-life, it is not commercially feasible to distribute N-13 ammonia from a cyclotron center to PET centers. Rb-82 PET imaging for myocardial perfusion is the first indication for PET imaging that has been approved by the FDA and is reimbursed by Medicare. However, Rb-82 generators are very expensive (mainly due to very few centers using it) and published data demonstrate no major advantage in sensitivity or accuracy compared with the widely available stress thallium myocardial perfusion scintigraphy. Rb-82 PET imaging has fallen out of favor in recent years.

History and current reimbursement policy of PET

While PET scans emerged about the same time as computed tomography (CT) scanners, ¹ clinical use of PET imaging has been delayed compared with CT. The main obstacle to its acceptance has been a lack of reimbursement for PET scans, although this situation has improved recently. In the early 1990s, some private insurance companies approved payment for PET imaging on a case-by-case basis for certain brain tumors and epilepsy focus localization. In March 1995, the Healthcare Finance Administration (HCFA) finally approved the first indication for PET imaging with Rb-82 for diagnostic imaging of patients with coronary artery disease. Coverage for the use of FDG to evaluate myocardial viability is still under review. In 1998, HCFA began reimbursement for FDG-PET scanning for the evaluation of indeterminate solitary pulmonary nodules and for initial staging of non-small-cell lung cancer. In 1999, HCFA added three new indications: detection and localization of colorectal cancer with rising carcinoembyronic antigen (CEA), staging/restaging with characterization of Hodgkin's disease and non-Hodgkin's lymphoma, and identification of metastases in melanoma ³⁸ (Table 1). These added indications have set the benchmark for private insurance companies to follow.

Why HCFA has authorized Medicare to reimburse PET imaging for some tumors and has excluded others is not clear. But the principle mechanism of tumor detection with FDG-PET imaging is clear, which is based on the fact that almost all malignant cells have increased glucose metabolism, increased protein synthesis, and uncontrolled cell proliferation. The most widely used radiopharmaceutical in clinical PET is FDG, a glucose analog. It is transported into cells like glucose and phosphorylated to FDG-6-phosphate by hexokinase as glucose. However, FDG-6-phosphate is polar and does not cross cell membranes, therefore it is trapped in cancer cells. FDG-6-phosphate can be dephosphorylated to FDG by glucose-6-phosphatase, but this process occurs very slowly in malignant cells because of a deficiency or lack of glucose-6-phosphatase. Trapped FDG serves as a tracer of glucose metabolism proportional to the rate of true glucose metabolism. Primarily, FDG-PET images are interpreted by visual analysis and/or semiquantitative analysis, such as standardized uptake value (SUV), which is defined as the activity in the region of interest (ROI) in uCi/mL divided by the patient's weight in kilograms and injected dose of FDG in mCi. In some scenarios, SUV is a very useful tool to discriminate benign from malignant process.

This article addresses current oncologic indications covered by HCFA to illustrate the usefulness of FDG-PET categories and addresses its strengths and limitations.

Solitary pulmonary nodule

Each year, 150,000 new cases of solitary pulmonary nodule (SPN) are diagnosed. Statistically, approximately 40% of these SPN cases are malignant. FDG-PET imaging has been very useful in the characterization of SPN, with an overall sensitivity of 95%, specificity of 88% and accuracy of 94%. ^4,8-11 Generally, malignant SPNs are hypermetabolic in utilizing glucose, resulting in focally intense FDG uptake within the nodules (figure 1). In contrast, most benign nodules (calcified granulomas or scars) are not highly FDG avid. Semiquantitative measurements are helpful in some borderline cases. Currently the metabolic threshold for pulmonary malignancy used by many experts is SUV 2.6 to 3.0. Although the specificity of FDG-PET imaging is relatively high (88%), false-negative results can occur, especially in bronchiole-alveolar cell carcinoma. ¹² Therefore, SPN, which is metabolically inactive, should be followed up by a CT scan or chest radiograph. False-positive results may be seen in granulomatous disease, such as tuberculosis or fungal infection, and sarcoiclosis. Therefore, a hypermetabolic (hot) SPN on FDG-PET scan should be followed by a tissue biopsy to confirm the diagnosis; some investigators recommend proceeding directly to a thoracotomy, as long as the patient's clinical presentation permits. FDG-PET imaging is the single best modality for SPN characterization compared with CT and commonly used Bayesian analysis criteria. ⁸

Staging non-small-cell lung cancer

Lung cancer is the most common and deadly cancer in both men and women. Non-small-cell lung cancer (NSmCLCa) accounts for 70% to 80% of all cases of lung cancer. Successful surgical resection with intent to cure NSmCLCa depends on the extent of the cancer. Preoperative staging of mediastinum with CT is only 50% accurate. Mediastinocopy has higher sensitivity than CT (approximately 85% to 90%), but is invasive and cannot access all thoracic station lymph nodes. Furthermore, up to 25% to 33% patients with lung cancer determined to be resectable by conventional imaging staging criteria are found at surgery to have unresectable disease. ⁴ This underestimation of extended disease is due to the fact that morphologic modalities are limitated in identifying metastatic lesions and in characterizing borderline-sized lymph nodes.

Multicenter studies have shown that FDG-PET imaging has very high sensitivity (88%) and specificity (93%) for staging NSmCLCa, which is comparable to that of mediastinoscopy (90% to 95%). FDG-PET imaging, however, has the advantages of lower cost, less morbidity, and accessibility to all station lymph nodes. It is particularly useful in high-risk surgical candidates, in whom there has been non-yield of biopsy or difficult biopsy site, and patients with bleeding disorders threatening biopsy attempts.

Small-cell lung cancer staging can be done equally well with FDG-PET imaging, but it is not reimbursed by HCFA since the majority of patients with small-cell lung cancer already have distant metastases at the time of presentation. Chemo/radiation therapy is the mainstream of the treatment. Staging small-cell lung cancer with PET may not alter patients' treatment.

Recurrent colorectal cancer

Colorectal cancer is the second leading cancer death overall in western industrialized countries. Primary treatment is surgical resection with an approximate 50% 5-year survival rate. Reportedly, 2% to 50% of patients who have undergone surgical resection develop local recurrence within the following 2 years. If recurrent cancer is detected earlier, re-resection may still lead to a improved survival rate. To detect recurrent colorectal cancer, the tumor marker such as CEA is seen in only active disease, but in <50% of cases. Also, elevated CEA does not help to localize the site of recurrence. Because of postoperative changes and altered local anatomic structures, morphologic modalities such as CT, MRI and ultrasound cannot differentiate postoperative scar from tumor recurrence until the lesion becomes obvious in late stages. A negative biopsy result cannot exclude early recurrence due to potential tissue sampling errors. Multiple reports ^4,5,13-21 have shown that FDG-PET scanning is highly accurate in detecting early localized tumor recurrence with about 95% sensitivity, 98% specificity, and 96% accuracy, while CT provides only 65% accuracy. FDG-PET scanning evaluates the chest, abdomen, and pelvis in one examination setting, which permits identification of local recurrence as well as distant metastases. FDG-PET is also highly sensitive in detecting hepatic and extrahepatic metastases with 90% sensitivity. In contrast, the immunoscinitgraphic imaging such as Oncoscint is only about 50% sensitive. ⁴

PET imaging has made substantial impact on patient clinical management. FDG-PET scanning can distinguish post-treatment (operative/ radiation therapy) scar from recurrent tumor since malignant tumors are metabolically active and FDG-avid on PET imaging but scar tissue is not (figure 3). This high accuracy in identifying early-stage recurrent tumors with FDG-PET is crucial for surgical cure and improving patient outcome. Current literature ^4,13-20 shows that PET scans often detect unsuspected metastases not seen on morphologic imaging such as CT, MRI, and ultrasound. When PET scanning is used as part of workup for suspected recurrent colorectal cancer, the findings alter surgical planning in 27% to 63% (average 38%) of cases, which results in cost savings and avoidance of unnecessary surgery.

Lymphomas

In the past, CT scans and gallium scinitgraphy were the standard modalities for accurate staging and follow-up for patients with lymphoma (Hodgkin's disease and non-Hodgkin's lymphoma). However, as in colorectal cancer, morphologic imaging modalities have obvious limitations in differentiating post-treatment scar tissue from residual or recurrent lymphoma. For instance, using CT scanning to evaluate lymphoma requires size criteria and enlargement of lymph nodes on serial CT scans to confirm malignant disease, which increases costs and takes months, delaying the initiation of potentially curative treatment. Gallium-67 has limitations in resolving small or low-grade lesions and is problematic in evaluating the abdomen, because physiologic uptakes within the liver and spleen hinder detection of liver lesions. FDG- PET can also replace bone scans in primary staging of malignant ymphoma. ³⁶

Accumulated data ^21-28 have demonstrated that FDG-PET is ideal modality to stage lymphoma and follow-up tumor treatment response, since hypermetabolic foci indicate active disease in a treated site, while the absence of intense of FDG uptake is typical of fibrosis. Several studies also show FDG-PET is superior to gallium-67 scintigraphy, i.e. to detect more non-Hodgkin's lesions than gallium-67 especially in low grade neoplasms. Most clinic series and our experience show that PET scans can detect nearly all untreated lymphomas. FDG uptake is higher in untreated high-grade lymphomas than in low-grade lymphomas; although there is some overlap in SUVs. Many centers use FDG-PET to stage lymphoma and to determine if lymphadenopathy seen on CT after treatment is metabolically active or inactive, and thus whether additional chemotherapy is required. The current trend confirms our previous statement: ¹ "If PET is available, gallium scinitgraphy and lymphangiography play little role in lymphoma evaluation." In some centers and in our institute, FDG-PET has successfully replaced gallium scintigraphy for evaluation of residual and recurrent lymphomas (figure 4).

Malignant melanoma

The incidence of malignant melanoma is rapidly increasing and its incidence has increased more than 5% per year since early 1970s, more than any other malignancy. Melanoma is curable only when diagnosed early with complete resection from skin. Early diagnosis depends on regular skin examination and excisional biopsy of suspicious lesions. Malignant melanoma has a high mortality secondary to nodal metastases and unpredictable metastatic behavior. Melanoma is one of the most metabolically active tumors and is FDG-avid, make it well suited for FDG-PET imaging. Several studies ^29-34 have shown that FDG-PET scanning has an obvious advantage over morphologic imaging modalities such as CT, MRI, and ultrasound in the early staging of the disease and and is more accurate in staging melanoma patients than conventional images. Steinert and associates ³¹ reported 33 cases of known melanoma scanned with FDG-PET with overall sensitivity 93% and specificity 100%. The accuracy for melanoma detection is 90% to 95% for FDG-PET versus 40% to 80% for conventional workups. Actually, FDG-PET detects more unsuspected lesions with one whole-body scan than multiple CTs, which reveals unresectable disease and prevents unnecessary surgery on patients who would not benefit from surgery due to extensive metastasis (figure 5). A cost effectiveness analysis shows that 36% of cases scheduled for surgical resection based on conventional images changed into medical treatment based on FDG-PET findings of previously unsuspected additional lesions. The cost saving was $2200 per patient and considerable morbidity and mortality associated with surgery were avoided.

Where do we go from here?

Because of the current relief of many regulatory challenges and new PET reimbursement policies, the availability of FDG-PET scanners has progressed rapidly nationwide. Currently, there are approximately PET centers in the United States. The current cost of a dedicated PET system ranges from $1 to $2.5 million dollars without an onsite cyclotron, the latter costs an additional $2 to $2.5 million dollars. New PET scanners have better spatial resolution and more efficient counting capacity as well as two- and three-dimensional acquisition modes. Mobile PET scanners are now also available commercially. Several commercial radiopharmaceutical companies have developed regional supply sites to distribute FDG to PET centers that cannot afford a cyclotron. Each FDG dose (about 10 mCi) costs about $600 to $700 and is ordered the evening before the imaging is performed and is delivered early the next morning. Medicare reimbursement for approved indications of FDG-PET imaging range from $1,700 to $2,500 depending on geographic regions, which includes the cost of the radiopharmaceutical, technique fee, and radiologists' professional fee.

Dual-head hybrid coincidence cameras (conventional dual detectors cameras coupled with high-energy collimator and coincidence detection circuitry) have gained popularity in recent years because they are less expensive than dedicated PET scanners. Although these cameras suffered originally from inferior count sensitivity and spatial resolution compared with dedicated PET scanners, their performance has continuously im-proved and they provide a more accessible and less expensive alternative for small hospitals/physician groups to offer metabolic imaging services. Published data indicate that dual-head coincidence cameras are significantly superior in tumor detection when compared with conventional imaging modalities such as CT. Good agreement between the results of a dual-head coincidence camera and dedicated PET scanner has been reported in detecting hypermetabolic lesions more than 10 to 15 mm in 70 patients with histologically proven malignant tumors, or suspected recurrent or metastatic disease. ³⁷

Other oncologic indications appear very promising. Currently, the following indications for FDG-PET scanning are under HFCA consideration: 1) detecting primary and recurring head and neck cancer; ^39,40 2) differentiating residual and recurrent brain tumors from post-treatment necrosis; ^1,4,5 3) detection and localization of intermediate breast lesions on mammograms or in patients with breast implantation, as well as evaluation of axillary lymph node metastases with sensitivity and specificity ranging from 80% to 100%; ^1,4,43-46 and 4) assessment of response to cancer treatment. ^1,42,47,48

In the end of 19th century, Dr. Roentgen discovered X-ray, which made medical imaging an exciting and challenging medical specialty. In the past decade, FDG-PET imaging had revolutionized tumor detection and characterization. At the dawn of the 21st century, the human gene will be completely decoded and modern medicine will focus on gene therapy, ^49,50 which demands functional and molecular imaging--this is what PET imaging is all about. Future wide utilization of FDG-PET imaging to improve patient care and corresponding better reimbursement policy are expected. Finally, we finally can say without hesitation, clinical PET imaging's time has arrived!

Acknowledgment

The authors would like to thank Marcia K. Tan for typing the manuscript, and Kiayona Torreu and Jim Doyle for their technical assistance.

References

1. Tan TXL, Pomeranz SJ: PET: A new frontier in diagnosis and patient care. Applied Radiology 24(12): 6-24, 1995.

2. Tan TXL, Spigos DG, Mueller CP: Abnormal cortical metabolism in disseminated encephalo-myelitis. Clin Nucl Med 23(9): 629-630, 1998.

3. Tan TXL, Pretorius HT, Pomeranz PS, et al: Effects of myocardial viability assessment with positron emission tomography on clinical management and patient outcomes. Chin Med J 109(9): 687-694, 1996.

4. Conti PS, Lifien DL, Hawley K, et al: PET and FI-FDG in oncology: A clinical update. Nucl Med Biol 23(6): 717-735, 1996.

5. Delbeke D: Oncological applications of FDG PET imaging: Brain tumors, colorectal cancer, lymphoma and melanoma. J Nucl Med 40(4): 591-603, 1999.

6. Lowe VJ, Naunheim KS: Positron emission tomography in lung cancer. Ann Thoracic Surg 65: 1821-1829, 1998.

7. Maron EM, McAdams HP, Erasmus JJ, et al: Staging non-small cell lung cancer with whole-body PET. Radiology 212(3):803-809, 1999.

8. Dewan NA, Shehan CJ, Reeb SD, et al: Likelihood of malignancy in a solitary pulmonary nodule. Comparison of Bayesian analysis and results of FDG-PET scan. Chest 112(2):416-422, 1997.

9. Patz EF, Lowe VJ, Hoffman JM, et al: Focal pulmonary abnormalities: evaluation with F-18 fluorodeoxyglucose PET scanning. Radiology 188:487-490, 1993.

10. Gupta NC, Frank AR, Dewan NA, et al: Solitary pulmonary nodules: Detection of malignancy with PET with 2-[F-18]-fluoro-deoxy-D-glucose. Radiology 184:441-444, 1992.

11. Coleman RE: Multicenter perspective study of FDG-PET in the evaluation of indeterminate solitary pulmonary nodules. J Nucl Med 36:94P, 1995.

12. Torzuka T, Zasadry KR, Wahl RL: Diabetes decreases FDG accumulation in primary lung cancer. Clin Pos Imag 2(5):281-287, 1999.

13. Lai DTM, Fulham M, Stephen MS, et al: The role of whole-body positron emission tomography with [ ¹⁸ F] fluorodeoxyglucose in identifying operable colorectal cancer metastases to the liver. Arch Surg 131:703-707, 1996.

14. Flanagan FL, Dehdashti F, Ogunbiyi OA, et al: Utility of FDG-PET for investigating unexplained plasma CEA elevation in patients with colorectal cancer. Ann Surg 227:319-323, 1998.

15. Delbeke D, Vitola JV, Sandier MP, et al: Staging recurrent metastatic colorectal carcinoma with PET. J Nucl Med 38:1196-2101, 1997.

16. Schiepens C, Penninckx F, DeVadder N, et al: Contribution of PET in the diagnosis in the recurrence of colorectal cancer: comparison with conventional imaging. Eur J Surg Oncol 21:517-522, 1995.

17. Abdel-Nabi H, Doerr RJ, Lamonica DM, et al: Staging of primary colorectal carcinomas with fluorine-18-fluorodeoxyglucose whole-body PET: Correlation with histopathologic and CT findings. Radiology 206:755-760, 1998.

18. Strause LG, Clorius JH, Schiag P, et al: Recurrence of colorectal tumors: PET evaluation. Radiology 170:329-332, 1989.

19. Vogel SB, Drane WE, Ross PR, et al: Prediction of surgical resectability in patients with hepatic colorectal metastases. Ann Surg 219:508-516, 1994.

20. Kim EE, Chung SK, Haynie TP, et al: Differentiation of residual recurrent tumors for post-treatment changes with F-18 FDG PET RadioGraphics 12:269-279, 1992.

21. Newman J, Francis IR, Kiminske MS, Wahl RL: Imaging of lymphoma with PET with 2-[F-18]-fluoro-2-deoxyglucose: Correlar with CT. Radiology 190:111-116, 1994.

22. Paul R: Comparison of fluorine-18-deoxyglucose and gallium-67 citrate imaging for detection of lymphoma. J Nucl Med 28:288-292, 1987.

23. Okada J, Yoshhikawa K, Imazeki K, et al: The use of FDG-PET in the detection and management of malignant lymphoma: Correlation of uptake with progress. J Nucl Med 32:686-691, 1991.

24. Lapela M, Leskinen S, Minn HR, et al: Increased glucose metabolism in untreated NHL: A study with positron emission tomography and fluorine-18-fluorodeoxyglucose. Blood 86:3522-3527, 1995.

25. Hoh CK, Glaspy J, Rosen P, et al: Whole-body FDG-PET imaging for study of Hodgkin's disease and lymphoma. J Nucl Med 38:343-348, 1997.

27. Carr R, Barrington SF, Madan B, et al: Detection of lymphoma in bone marrow by whole-body by positron emission tomography. Blood 91: 3340-3346, 1998.

28. Rodriguez M, Rehn S, Ahistrom H, et al: Predicting malignancy grade with PET in NHL. J Nucl Med 36:1790-1796, 1995.

29. Gritter LS, Francis IR, Zasadry KR, Wahl RL: Initial assessment of positron emission tomography using 2-fluoro-18-fluoro-2-deoxy-D-glucose in the imaging of malignant melanoma. J Nucl Med 34: 1420-1427, 1993.

30. Domina DL, Fulham MJ, Thompson E, et al: Positron emission tomography in the detection and management of metastatic melanoma. Melanoma Res 325-329, 1996.

31. Steinert HC, Huch-Boni RA, Buck A, et al: Malignant melanoma: staging with whole-body positron emission tomography L 2-[F-18]-fluoro-2-deoxy-D-glucose. Radiology 195:705-709, 1995.

32. Rinne D, Baum RP, Hor G, et al: Primary staging and follup-up of high risk melanoma patients with whole-body IF-fluorodeoxyglucose positron emmision tomography: Results of prospective study with 100 patients. Cancer: 82:1664-1671, 1998.