Objectives

1. Review the biology of positron emission tomography (PET) scanning.
2. Understand the indications for and limitations of PET scanning in the evaluation of the solitary pulmonary nodule.
3. Review the utility of PET imaging vs traditional techniques in the staging of known lung cancer.
4. Understand the importance of appropriate patient selection for the application of this technology.
5. Review the current literature on the use of PET scan in the posttreatment follow-up of patients with lung cancer.

Abbreviations

FDG = fluorodeoxyglucose; PET = positron emission tomography; SPN = solitary pulmonary nodule; SUV = standardized uptake value

Lung cancer is the most common cause of cancer mortality in the United States. The majority of new cases are diagnosed at an advanced stage with attendant poor prognosis. The 5-year survival rate from the time of lung cancer diagnosis is 10 to 15%. The widespread availability of cross-sectional imaging has led to an increasing identification of lung nodules of uncertain biologic potential. This poses a clinical dilemma as physicians must struggle with the decision to refer patients for invasive and morbid diagnostic testing, lest they risk failing to diagnose this deadly cancer at a curable stage.

Current clinical staging modalities for lung cancer are plagued by both false-positive and false-negative evaluations of regional and mediastinal lymph nodes. Although surgery is clearly the procedure of choice (with or without adjuvant chemotherapy) in patients with N0 or N1 disease, it is used much less frequently and with much less success in those with ipsilateral or contralateral mediastinal disease (N2 or N3). CT often demonstrates lymph nodes that are enlarged for reasons other than nodal metastases, such as inflammation related to postobstructive changes or concomitant granulomatous disease. Conversely, it is not unusual for lymph nodes that are normal in size to harbor malignant cells at the time of surgical staging. PET scanning has been advocated as a means of improving the accuracy of clinical staging of lung cancer and allowing for better discrimination of those patients requiring mediastinoscopy for accurate prethoracotomy staging.

This review will briefly discuss the theory behind PET as well as its role in the evaluation of the solitary pulmonary nodule (SPN), the staging of known lung cancer, and the posttreatment surveillance of lung cancer patients.

Biology of PET Scanning

Lung cancers demonstrate increased uptake of glucose relative to normal tissues. In addition, tumor cells frequently have elevated levels of hexokinase and decreased levels of glucose-6-phosphatase, which results in increased conversion of glucose to glucose-6-phosphate, and a decrease in the reverse reaction. Fluorodeoxyglucose (FDG) is a positron-emitting glucose analog that is transported into cells and phosphorylated by hexokinase (Fig 1). It is then unable to be metabolized further by the glycolytic pathway. In combination with decreased glucose-6-phophatase activity, this leads to accumulation of phosphorylated FDG in tumor cells.¹

PET scanning allows for the measurement of annihilation photons, which are produced by the collision of positrons –tiny, positively charged particles– with electrons, their negatively charged counterparts. Each collision produces two annihilation photons, which travel in opposite directions. The accumulation of 18-FDG in tumor cells and the measurement of the resulting annihilation photons from different angles allows for a three-dimensional assessment of FDG uptake, and thus metabolic activity in the region of interest.² Quantification of the test result using the standardized uptake value (SUV) provides a measure of the intensity of photon production as a function of the dose of 18-FDG administered and the patient's body mass.

PET Scanning in the Evaluation of the SPN

An SPN is a lesion up to 3 cm in size that is entirely surrounded by lung parenchyma. The malignant potential of an SPN varies substantially with the characteristics of the patient. It is estimated that 30 to 50% of SPNs are malignant, with the majority of the remainder representing benign granulomata. Age, smoking history, nodule size, and absence of calcification are all associated with an increased probability of malignancy.³ CT scanning is frequently indeterminate in differentiating benign from malignant nodules. Choosing the appropriate course of watchful waiting or aggressive early intervention is a frequent dilemma facing pulmonologists.

There has been a great deal of investigation into the role of PET scan in the evaluation of the SPN (Fig 2). Studies vary with regard to case definition, sample size, requirement for histologic confirmation of diagnosis, prevalence of malignancy in the study population, and the use of qualitative or semiquantitative techniques for determining PET results. This variability in study design has left the precise role of PET scanning open to interpretation.

Despite the shortcomings of the available literature, there are clearly settings in which PET imaging is more likely to be inaccurate. The spatial resolution of most PET scanners varies between 7 and 8 mm, increasing the false-negative rate for lesions <1 cm. Recent investigations using respiratory gating during image acquisition to reduce the "smearing" of images caused by respiratory motion may enhance the resolution of PET scans for smaller lesions by more accurately assigning the photons produced within the appropriate anatomic area.⁴ Larger prospective studies are necessary before this approach is validated. Hyperglycemia, particularly acute hyperglycemia, can decrease cellular uptake of FDG, leading to false-negative results.⁵ This is particularly concerning as diabetic patients are frequently asked to hold or reduce the dosage of their medication prior to diagnostic testing, and patients are generally asked to fast for several hours before PET scanning. Most centers require glucose levels <200 mg/dL to reduce the chances of a false-negative scan. Bronchoalveolar carcinoma and carcinoid tumors are associated with a higher false-negative rate, and PET scans should be interpreted with caution in settings where either of these entities is felt to be likely.^6,7

Active infectious processes often demonstrate increased FDG uptake, and the use of standard PET imaging is of dubious benefit when active infection is likely. Procuring images at more than one time point may allow for more accurate differentiation of benign and malignant lesions. Tumors tend to demonstrate increasing SUVs over time as many tumors will continue to take up FDG for several hours after injection, whereas benign lesions tend to plateau or decrease on serial imaging. Allowing a time lapse of at least 30 min between serial images allows for a larger change in SUV. In a study of 36 patients, the use of dual-time imaging was used in an effort to determine if differences in FDG kinetics would allow improved differentiation of benign from malignant lesions. The authors found that a prospectively defined increase of 10% in SUV between the first and second scan improved sensitivity from 80 to 100% in the 20 patients with malignant lesions. There were two patients with benign lesions who demonstrated an increase in SUV of >10%, but both had SUVs <1.0. The authors suggest that using dual-time imaging for lesions with SUV >1.0 might allow for improved detection of malignant lesions.⁸ Larger-scale studies are warranted before this technique can be recommended as the standard of care.

A recent meta analysis examined the role of PET scanning in the evaluation of the SPN.⁹ The authors compared the use of SUV-guided interpretation of images with qualitative interpretation and assessed the usefulness of PET scans in the evaluation of pulmonary nodules of varying sizes. The authors included studies that evaluated at least 10 patients with an SPN or lung mass, in which at least five cases proved to be malignant, and that presented sufficient data to allow calculation of sensitivity and specificity. A total of 40 studies were included with 1,474 focal pulmonary lesions. Sensitivity ranged from 83 to 100%, with wider ranges of specificity. The authors found that PET scanning in current practice operates at a sensitivity of 96.8% and a specificity of 77.8%. There appeared to be no difference between the results obtained using an SUV cutoff of 2.5 compared with a qualitative visual assessment for increased uptake. Whether this will remain the case with the use of dual-time point imaging remains to be seen. In addition, the accuracy of PET scan was not significantly different when comparing nodules <3 cm with lesions of any size. Though there were only eight lesions <1 cm evaluated across these studies, three of the six malignant lesions were misclassified as benign.

Appropriate patient selection is of paramount importance even in light of the high sensitivity afforded by PET scanning. In those with a high likelihood of malignancy, the finding of a negative PET scan is still associated with a significant likelihood of malignancy. Using the likelihood ratios generated by Gould et al,⁹ a patient with a pretest probability of malignancy of 80% still has a 14% posttest probability of cancer with a negative PET scan. In this patient population, PET scanning is unlikely to alter the need for a tissue diagnosis, and is not warranted in the majority of such cases. Similarly, PET is of dubious benefit in patients with an extremely low likelihood of malignancy, given the poor positive predictive value in the setting.¹⁰ In this setting, serial CT seems a more prudent option, with biopsy reserved for those with a lesion enlarging over time. In those patients whose likelihood of malignancy is neither very low nor very high, PET scan is an extremely useful tool in determining which patients to refer for biopsy or surgery. Negative findings on PET in this setting substantially reduce the likelihood of lung cancer and make serial follow-up a more palatable option.

PET in the Staging of Known Lung Cancer

Once the diagnosis of lung cancer is made, decisions on the optimal management strategy hinge on accurate staging. The presence of malignant airway obstruction, concomitant infections, or unrelated processes can lead to enlargement of regional and mediastinal lymph nodes, which can be confused with metastatic disease. In addition, malignant cells may reside in lymph nodes of normal size, causing the clinician to underestimate the extent of disease. The ability to provide information about metabolic activity in lymph nodes of normal and abnormal size has led to the investigation of PET as an imaging procedure that is complementary or alternative to CT (Fig 3). The available literature is limited by small sample sizes, variable result reporting (patient-based vs lymph node-based reporting), and differing prevalence of mediastinal metastases.

Gould et al¹¹ published a meta analysis of studies comparing FDG-PET and CT scan in the mediastinal staging of non-small cell carcinoma. They examined the performance characteristics of PET and CT separately as well as the impact of CT findings (the presence or absence of lymph nodes >1.0 cm) on the performance of PET. A total of 39 studies evaluating 3,110 patients were included. The authors found that PET was more accurate than CT for staging of the mediastinum. Using the summary receiver operating characteristic curves generated from these 39 studies, the authors estimated that the sensitivity of CT is 59% in current practice, while that of PET is 81%. Specificity also favors PET, with CT-based staging having 79% specificity vs 90% for PET-based staging. As one might expect from the data on SPNs, PET was found to operate on a different portion of the receiver operating characteristic curve in the presence of enlarged lymph nodes on CT. When CT revealed enlarged lymph nodes, the median sensitivity of PET was 91%, with a median specificity of 78%. In the absence of enlarged nodes on CT, the sensitivity of PET decreased to 75%, with a median specificity of 93%.

Although PET is more accurate than CT in the staging of the mediastinum in non-small cell lung cancer, its performance is not sufficient to obviate the need for invasive lymph node sampling in the presence of a positive PET scan. The implications of a false-positive PET scan on the management strategy used and the probability of cure dictate that positive PET scans require confirmatory lymph node biopsy except in settings where the pretest probability of disease is extraordinarily high or the underlying condition of the patient precludes it. In the setting of a negative PET scan, the use of mediastinoscopy or alternative lymph node sampling is less clear. In the presence of enlarged mediastinal lymph nodes and an estimated pretest probability of mediastinal metastasis of 35%, a negative PET is associated with a posttest probability of mediastinal disease of 13%. In the absence of lymph node enlargement, the posttest probability of mediastinal metastasis in the presence of a negative PET is 9%.¹¹ Bypassing mediastinoscopy and proceeding to thoracotomy in the presence of a negative PET scan seems reasonable for patients with T1 lesions in whom the probability of mediastinal involvement is lower, but the decision should be individualized for patients with locally advanced or central disease.¹²

FDG-PET also offers the ability to detect otherwise occult metastatic disease. Studies have demonstrated the presence of metastatic disease in 11 to 24% of patients undergoing PET scan for staging of lung cancer.^13,14 Whether these metastases were or should have been clinically suspected cannot be determined by reviewing these studies. A meta-analysis by Hellwig et al¹⁵ revealed extrathoracic metastases in 12% of 581 evaluable cases and estimated that the management course was changed in 18% of patients who underwent PET scan for lung cancer staging. Bone scintigraphy has excellent sensitivity for detecting skeletal metastases, but has been limited by poor specificity and positive predictive value. In a prospective evaluation, 48 patients with non-small cell cancer were evaluated with bone scintigraphy and PET scan. Diagnosis of bony metastasis was confirmed by pathology or progressive disease over 1 year of radiographic follow-up. A total of 138 bony lesions were identified, with 106 representing metastatic lesions and 32 representing benign disease. PET scan correctly identified 99 of the 106 malignant lesions and 30 of the 32 benign lesions, compared with 98 and 2, respectively, for bone scintigraphy. The diagnostic accuracy of PET scan was 92.5% as compared with 72.5% for bone scintigraphy.¹⁶ Given its ability to detect mediastinal disease and distant metastases in sites other than the skeleton, it seems reasonable to replace bone scintigraphy with FDG-PET imaging in the evaluation for skeletal metastases. It must be noted, however, that PET cannot be relied upon to accurately assess for intracranial disease using present techniques owing to the high glucose avidity of brain tissue.

Recent attention has been paid to the use of integrated PET and CT in the staging of lung cancer. The less precise morphologic information provided by PET scan as compared with CT can make the precise location of abnormal lesions difficult to identify. This has the potential to misclassify N1 nodes as N2 or vice versa . PET-CT scanners allow sequential acquisition of CT and PET images to be taken by scanners aligned in series, with the patient remaining on the same examining table. The images are coregistered on the same hardware, allowing direct comparison of anatomic and functional lesions. A prospective comparison of integrated PET-CT with CT alone, PET alone, and visual correlation of PET and CT in 50 patients demonstrated improved accuracy of lymph node staging, and improved anatomic localization of extrathoracic metastases using integrated PET-CT when compared with CT alone, PET alone, or visual correlation of PET and CT.¹⁷ The impact on patient management of a strategy employing integrated PET-CT has yet to be determined.

The Use of PET in the Posttreatment Follow-up of Lung Cancer

An accurate assessment of the response to treatment (Fig 4) could be of benefit in providing information on prognosis or guiding the aggressiveness of future therapy. Knowledge of the degree of residual disease posttreatment is also important in those patients with advanced disease who are being considered for surgical therapy. In stage IIIA non-small cell lung cancer, neoadjuvant chemotherapy or chemoradiotherapy is sometimes employed in an effort to down-stage disease to allow surgical excision. In patients who are successfully down-staged by neoadjuvant therapy from N2 to N0 disease, 5-year survival may exceed 35%, but remains <10% for those with residual N1 or N2 disease after therapy.¹⁸ Most patients in whom surgical therapy is planned for advanced-stage disease will undergo surgical restaging of the mediastinum prior to definitive surgical treatment. A second mediastinoscopy, particularly after radiation therapy, is technically challenging. A noninvasive imaging study that could provide information comparable to that obtained by mediastinoscopy would certainly have a role in the management of these patients.

The usefulness of FDG-PET in the setting of neoadjuvant therapy has been the topic of several studies. The Leuven Lung Cancer Group performed a pilot study evaluating the use of PET in 15 patients with pathologic N2 disease. PET scans were performed prior to therapy and 4 weeks after therapy. Of the 15 patients, nine went on to surgery after neoadjuvant therapy. The PET scan was accurate in all of the patients who were restaged surgically, with a sensitivity and specificity of 100%. Of the total group, seven had persistent activity in the mediastinum posttherapy; all of these patients suffered early relapse and death. The remaining eight patients had a negative posttreatment PET scan. Six of these patients remained alive and free of disease at the time of reporting (mean follow-up, 15.9 months), one had suffered early relapse and death, and one had isolated intracranial metastasis treated with radiation therapy.¹⁹

More recent investigations have been mixed in their results. Cerfolio et al²⁰ published their results in 34 patients treated with neoadjuvant chemotherapy or chemoradiotherapy. A PET scan was performed before and a median of 23 days after therapy was completed. PET scan was evaluated in comparison to CT and pathologic stage for the primary tumor, N1 nodes, and N2 nodes (the authors divided N2 nodes into two groups based on accessibility by mediastinoscopy). PET attained high sensitivity (97%) in the evaluation of the primary tumor; the specificity was 67%. Overall accuracy for the assessment of the primary tumor was 94%. PET performed poorly in the evaluation of N1 nodes, with only 18% sensitivity and 95% specificity. In the evaluation of N2 disease, sensitivity of PET was 73% and specificity was 100%. For evaluation of the primary tumor and N2 lymph nodes, PET was superior to CT. Port et al²¹ reported their results in 25 patients enrolled in two phase II trials of induction therapy for stage IB to IIIA non-small cell cancer. PET scan was performed prior to therapy and 2 weeks after therapy. A PET response was defined as a reduction in SUV of at least 50%. In this study, the results of PET were disappointing, with only 68% accuracy when predicting a major pathologic response at the site of the primary tumor and only 48% accuracy in restaging mediastinal disease, with a sensitivity of only 61% for residual disease. The reason for the disparate results may relate to the timing of PET after completion of therapy, with increasing accuracy expected as the inflammatory changes induced by therapy abate. Further investigation is clearly needed before restaging by means other than mediastinoscopy can be recommended.

Evaluation of response to treatment may also be of value in those patients who are not being considered for surgical therapy as it may provide information about prognosis or guide additional treatment strategies. The lack of a looming operation allows a longer period prior to rescanning, which may improve the accuracy of PET. Bury et al²² evaluated response to therapy with PET and CT in 126 patients with non-small cell cancer. PET scans were performed within 6 months of completion of therapy and every 6 months thereafter for a follow-up period of 8 to 40 months. PET exhibited 100% sensitivity for detection of recurrence in the 60 patients with recurrent disease, with a specificity of 92%. Mac Manus et al²³ reported their results in 73 patients evaluated with CT and PET before therapy and a median of 70 days after treatment. Posttreatment scans were classified as revealing complete response (no residual uptake), partial response (significant decrease in uptake with no new sites of disease), no response (no change), or progressive disease (increased uptake or new sites of disease). There was poor correlation between PET and CT, with PET revealing more patients with complete response and serving as the only predictor of survival in multivariate analysis. One-year survival for those with complete response compared with partial response or no response on PET scan was 89% vs 60% vs 27%. Likewise, Patz et al²⁴ demonstrated that PET scan powerfully predicted prognosis after initial therapy including either surgery, radiotherapy, or chemotherapy for non-small cell cancer in their cohort of 113 patients. One hundred patients with a positive PET scan posttherapy had a median survival of 12 months, whereas 11 of 13 patients with a negative PET posttherapy were alive at the end of the 4 to 60 months of follow-up. The two patients who died lived for 28.5 and 43.1 months after treatment, respectively. PET scans were performed a median of 8 months after therapy.

Summary

FDG-PET scanning allows for an assessment of the metabolic activity of tissue, which facilitates improved discrimination of benign lung nodules from malignant ones. In the evaluation of the SPN, PET is of little use in the setting of a high or very low pretest probability of cancer, but can be used to justify a strategy of serial follow-up in patients with a low or intermediate probability of cancer. Techniques, such as dual-time point imaging and respiratory gating, may result in enhanced performance of PET in the future. PET provides superior accuracy compared with CT in the mediastinal staging of known non-small cell lung carcinoma. Given the imperfect specificity, all positive PET scans should be followed by pathologic confirmation to avoid incorrectly precluding a patient from surgery. The available literature is conflicting with regard to the usefulness of PET for restaging the mediastinum in patients with advanced disease (stage IIIA) prior to possible surgical resection. This may relate to the short time interval between treatment and follow-up PET, but larger studies are needed before the role of PET, if there is one, is clear in this setting. In contrast, PET provides useful information and is superior to CT in evaluating response to therapy in patients who are not treated surgically.

Acknowledgment

References

1. Gallagher BM, Fowler JS, Gutterson NI, et al. Metabolic trapping as a principle of radiopharmaceutical design: some factors responsible for the biodistribution of [18F] 2-deoxy-2-fluoro-D-glucose. J Nucl Med 1978; 19:1154-1161  
2. Mac Manus MP, Hicks RJ. PET scanning in lung cancer: current status and future directions. Semin Surg Oncol 2003; 21:149-155  
3. Lillington GA. Management of solitary pulmonary nodules. Dis Mon 1991; 5:271-318  
4. Nehmeh SA, Erdi YE, Ling CC, et al. Effect of respiratory gating on quantifying PET imaging of lung cancer. J Nucl Med 2002; 43:876-881  
5. Lowe VJ, Naunheim KS. Current role of positron emission tomography in thoracic oncology. Thorax 1998; 53:703-712  
6. Higashi K, Ueda Y, Seki H, et al. Fluorine-18-FDG PET imaging is negative in bronchoalveolar cell carcinoma. J Nucl Med 1998; 39:1016-1020  
7. Erasmus JJ, McAdams HP, Patz EF Jr, et al. Evaluation of primary pulmonary carcinoid tumors using FDG PET. AJR Am J Roentgenol 1998; 170:1369-1373  
8. Matthies A, Hickeson M, Cuchiara A, et al. Dual time point 18F-FDG PET for the evaluation of pulmonary nodules. J Nucl Med 2002; 43:871-875  
9. Gould MK, Maclean CC, Kuschner WG, et al. Accuracy of positron emission tomography for diagnosis of pulmonary nodules and mass lesions: a meta-analysis. JAMA 2001; 285:914-924  
10. Detterbeck FC, Falen S, Rivera MP, et al. Seeking a home for a PET, part 1: defining the appropriate place for positron emission tomography imaging in the diagnosis of pulmonary nodules or masses. Chest 2004; 125:2294-2299  
11. Gould MK, Kuschner WG, Rydzak CE, et al. Test performance of emission tomography and computed tomography for mediastinal staging in patients with non-small-cell lung cancer: a meta-analysis. Ann Intern Med 2003; 139:879-892  
12. Detterbeck FC, Falen S, Rivera MP, et al. Seeking a home for a PET, part 2: defining the appropriate place for positrion emission tomography imaging in the staging of patients with suspected lung cancer. Chest 2004; 125:2300-2308  
13. Gupta NC, Graeber GM, Rogers JS 2nd, et al. Comparative efficacy of positron emission tomography with FDG and computed tomographic scanning in preoperative staging of non-small cell lung cancer. Ann Surg 1999; 229:286-291  
14. Kalff V, Hicks RJ, MacManus MP, et al. Clinical impact of 18 F fluorodeoxyglucose positron emission tomography in patients with non-small cell lung cancer: a prospective study. J Clin Oncol 2001; 19:111-118  
15. Hellwig D, Ukena D, Paulsen F, et al. [Meta-analysis of the efficacy of positron emission tomography with F-18-fluorodeoxyglucose in lung tumors: basis for discussion of the German Consensus Conference on PET in Oncology 2000]. Pneumologie 2001; 55:367-377  
16. Hsia TC, Shen YY, Yen RF, et al. Comparing whole body 18F-2-deoxyglucose positron emission tomography and technetium-99m methylene diophosphate bone scan to detect bone metastases in patients with non-small cell lung cancer. Neoplasma 2002; 49:267-271  
17. Lardinois D, Weder W, Hany TF, et al. Staging of non-small-cell lung cancer with integrated positron-emission tomography and computed tomography. N Engl J Med 2003; 348:2500-2507  
18. Bueno R, Richards WG, Swanson SJ, et al. Nodal stage after induction therapy for stage IIIA lung cancer determines patient survival. Ann Thorac Surg 2000; 70:1826-1831  
19. Vansteenkiste JF, Stroobants SG, De Leyn PR, et al. Potential use of FDG-PET scan after induction chemotherapy in surgically staged IIIa-N2 non-small-cell lung cancer: a prospective pilot study; the Leuven Lung Cancer Group. Ann Oncol 1998; 9:1193-1198  
20. Cerfolio R, Ojha B, Mukherjee S, et al. Positron emission tomography scanning with 2-fluoro-2-deoxy-d-glucose as a predictor of response of neoadjuvant treatment for non-small cell carcinoma. J Thorac Cardiovasc Surg 2003; 125:938-944  
21. Port JL, Kent MS, Korst RJ, et al. Positron emission tomography scanning poorly predicts response to preoperative chemotherapy in non-small cell lung cancer. Ann Thorac Surg 2004; 77:254-259  
22. Bury T, Corhay JL, Duysinx B, et al. Value of FDG-PET in detecting residual or recurrent nonsmall cell lung cancer. Eur Respir J 1999; 14:1376-1380  
23. Mac Manus MP, Hicks RJ, Matthews JP, et al. Positron emission tomography is superior to computed tomography scanning for response-assessment after radical radiotherapy or chemoradiotherapy in patients with non-small-cell lung cancer. J Clin Oncol 2003; 21:1285-1292
24. Patz EF, Connolly J, Herndon J. Prognostic value of thoracic FDG PET imaging after treatment for non-small cell lung cancer. AJR Am J Roentgenol 2000; 174:769-774