PET Imaging in Bronchogenic Carcinoma:

General:

The PET agent 2-(fluorine-18) fluo-2-deoxy-D-glucose (18-FDG) has been the most studied for evaluation of bronchogenic carcinoma. 18-FDG competes with glucose for facilitated transport into tumor cells and also competes with glucose for phosphorylation by hexokinase. Unlike glucose, however, the phosphorylated form is not further metabolized, it becomes trapped within the cell with little back diffusion or degradation. Uptake of FDG therefore reflects regional rates of exogenous glucose utilization. The agent is useful for imaging bronchogenic carcinoma because lung tumor cells have an increased cellular uptake of glucose due to an increased number of surface transport proteins and a higher rate of glycolysis in comparison to non-neoplastic cells (ie: increased metabolic rate in lung cancer cells). FDG uptake is related to the degree of cell differentiation- increased Glut-1 glucose transporter expression, and hence increased FDG uptake, is associated with greater degrees of de-differentiation [1]. Finally, FDG uptake is also related to the proliferation potential of lung cancer as estimated by the proliferating cell nuclear antigen labeling index [1]

FDG PET imaging can have a significant impact on patient management by heightening suspicion for pulmonary malignancy, identifying unsuspected sites of disease, and by guiding selection of a biopsy site. Similarly, a negative PET scan indicates a low likelihood for malignancy and supports the use of conservative management and follow-up [40]. In a study evaluating the impact of FDG PET exam results in the clinical management of patients with suspected or known thoracic malignancies PET scans influenced treatment in 65% of patients and offered new information in 85% [40].

Solitary Pulmonary Nodule Evaluation:

The differential for a solitary pulmonary nodule (SPN) is quite broad, but a large percentage of these lesions may prove to be malignant (especially if the lesion is larger than 2 cm). In the evaluation of an SPN, only 10-20% of patients with malignancy will have sputum positive for cancer, and up to 30% of patients with negative transthoracic biopsies are ultimately found to have malignancy [5]. FDG PET imaging can be used to evaluate indeterminate solitary pulmonary nodules to determine whether the nodule is benign or malignant. A standardized uptake ratio (SUR) is used to determine if a lesion has increased 18-FDG activity. The SUR normalizes the amount of FDG accumulation in a region of interest (ROI) to the total injected dose and the patient's body weight. It provides a means of comparison of FDG uptake between patients [41]. The SUR is calculated by dividing the mean activity within a selected region of interest (in mCi/ml) by the injected dose (in mCi/kg). Modifications of the SUR that may improve the semiquantitative evaluation of FDG uptake include using body surface area or the lean body weight instead of the weight of the patient; this is significant because the distribution of FDG is higher in muscle than in fat [44]. It is also important to remember that SUR values may change with time after FDG injection; thus, the time of acquisition after FDG injection must be standardized for the values to be useful [25]. Other factors which can affect the SUV include partial-volume effects, patient motion (due to lesion blurring), and the blood glucose level at the time of injection [46,53].

SUR= Mean selected region activity (mCi/ml)/injected dose (mCi)/body weight (kg)

An SUR greater than 2.5 has been shown to be very sensitive and specific for malignant lesions [13]. Visual analysis of the amount of uptake within a lesion has also been shown to be effective in differentiating benign from malignant lesions [13,47]. Uptake greater than blood pool activity (i.e.: liver or mediastinum typically have an SUV of 2.0) indicates a malignant lesion, while activity equal to or less than mediastinal blood pool suggests a benign lesion [47]. In fact, visual analysis may be more sensitive for nodules smaller than 1.5 cm in size (although, the improved sensitivity comes at a decreased specificity) [47]. 

High sensitivities (82-100%), specificities (67-100%), and accuracy (79-94%) have been reported for the identification of pulmonary malignancy using FDG PET imaging [25,47,50,52]. However, the high incidence of cancer in the many of the study groups may contribute to the high reported sensitivities. In a prospective multicenter study for the evaluation of pulmonary nodules (larger than 7 mm) felt to be indeterminate for malignancy based upon their conventional imaging appearance, FDG PET had a sensitivity of 92% and a specificity of 90% using SUR data [47]. Visual analysis of the images demonstrated a sensitivity of 98%, but a lower specificity of 69% [47]. Improved measurement of lesion volume and quantification of FDG uptake can be achieved with respiratory gated PET imaging [54]. Respiratory motion produces lesion blurring and an apparent increase in lesion size [54]. This results in a decrease in the activity concentration per pixel within the lesion and an underestimation of SUV [54]. The degree of lesion motion is dependent on location within the lung (apical lesions will move little, while lower lung lesions can move considerably) [54].

Dual time imaging:

Improved detection of malignant lesions can be obtained by performing dual time point imaging [16,53]. This technique takes advantage of the fact that tracer activity will washout of inflammatory lesions, while malignant lesions retain or increase in FDG activity over time [16,53]. In the study by Matthies et al [53], sensitivity and specificity of a standard SUV measurement greater than 2.5 for malignancy were 80% and 94%, respectively. For dual time imaging, using a threshold of a 10% increase in SUV between the two exams, sensitivity increased to 100% and specificity was 89% [53]. Lesions with a baseline SUV of less than 1.0 have a very high likelihood of being benign and dual time imaging is not required as it may result in false positive exams [53]. A limitation to consider when performing dual time imaging is patient motion which can significantly affect SUV determination, especially for low SUV values. Also- the change in SUV is dependent on reproducibility of the same ROI between both exams [53].

False positive results are not uncommon and can be seen with infectious granulomas (tuberculosis, histoplasmosis, aspergillosis, cryptococcus, and inflammatory pseudotumor) [36,50] or inflammatory processes (sarcoid, Wegener's), rheumatoid nodules [13,15], and an aggressive neurofibroma. [8] False positive lung nodule exams are thought to be related to increased glycolytic activity within activated macrophages [3].

Although uncommon, false negative exams can occur in three settings [41]:

1- A neoplasm with low metabolic activity: Bronchoalveolar cell carcinoma (BAC) may have a lower uptake value and can cause a false-negative result in up to 57% of cases [18,21,47,50]. Interestingly, Thallium-201 imaging is more commonly positive for BAC [1]. Another pulmonary neoplasm which frequently has a negative FDG PET exam (SUR < 2.5) is a carcinoid tumor (up to 85% of cases) [42,43,50]. Although less common, other NSCLC's such as squamous cell carcinoma and adenocarcinoma may also fail to demonstrate significant FDG accumulation [38,47]

2- Small lesions (under 1 cm to 1.5 cm in size) or lesions with a physically small histologic complement of malignant cells: Lesions as small as 1 cm can be accurately evaluated, but the evaluation of smaller lesions may result in false negative results due to spatial resolution limitations of PET imaging (approximately 5 mm) and respiratory motion [7,13,21]. FDG PET has shown a decreased sensitivity (80%) for malignancy for lesions smaller than 1.5 cm [41,47]. The ability to characterize smaller lesions is in part dependent on the degree of tracer accumulation within the lesion. Even very small lesions can be correctly characterized if they accumulate enough FDG to become visible. Sensitivity for lesions between 5-8 mm in size is between 50% to 73% [7,41].

3- Hyperglycemia [47]: Competitive inhibition from high serum glucose levels appears to hinder FDG uptake in some cases. This effect is most important with acute hyperglycemia while a chronically raised glucose level inhibits tumor uptake minimally (about 10%) [41].

Based upon the present data available- a lesion that is hot on PET should be considered malignant until proven otherwise. A lesion that is PET negative has a low probability for being a malignancy (under 5% [37]). However, one must consider all characteristics of a lesion prior to discounting its malignant potential. Follow-up exams should be performed on PET negative nodules to ensure stability [37]. If the lesion grows, further evaluation with tissue diagnosis should be obtained [37]. Patients who have a negative FDG PET exam but are subsequently shown to have lung cancer may have an overall better survival compared to patients with positive FDG exams [7].

Nuclear medicine gamma cameras equipped with ultrahigh energy collimators have also been used to perform SPECT images of 18-FDG in patients with suspected lung cancer. Overall sensitivity of SPECT FDG is 77% to 81% for identifying malignant pulmonary lesions (sensitivity has been report to be 100% for lesions larger than 2 cm [24]). Specificity is 91-100%. Small malignancies (less than 2 cm in size) are not reliably characterized with this imaging technique (sensitivity for lesions less than 2 cm is only 20-50%). The small number of patients studied and the high incidence of malignancy in the study groups (60-75%) make any conclusions about the true sensitivity of SPECT 18-FDG premature. [10]

Nuclear medicine dual detector gamma cameras have also been modified to permit coincidence detection. In general, these cameras have a limited intrinsic detector efficiency for the high energy FDG emissions which results in poor counting statistics. Non-true coincidence detection also contributes to image noise. None-the-less, pulmonary lesions larger than 1 cm seem to be well detected due to the relatively low background activity of the lungs. Lesions within the deep mediastinum and abdomen less than 1.5 cm in size, are less well identified [19]. However, despite technical improvements lesion detectability using hybrid gamma cameras is still limited compared to dedicated PET scanners [4].

Response to therapy:

An accurate assessment of the efficacy of chemotherapy and radiation therapy would prove of enormous value in directing therapy for patients with advanced stage NSCLC [45]. PET FDG imaging can evaluate the effectiveness of treatment and can also provide prognostic information regarding patient outcome following initiation of therapy [29].

Normalization of FDG uptake (i.e.: no uptake) after treatment indicates a good response to therapy and appears to be a good prognostic indicator [41,45]. Unfortunately, a decrease in FDG uptake may only indicate a partial response to treatment resulting from destruction of sensitive cells, while resistant cells continue to be metabolically active [41,45]. Preliminary data indicate that patients with positive PET scans following initial lung cancer therapy have a significantly worse prognosis than patients with negative results [11,29].  In one study, all patients with negative PET exams following treatment were alive at two years (even in the presence of non-specific radiologic changes), while 50% of patients with residual hypermetabolism had died during the same two year period [41,45]. Further evaluation will be required to determine if changes in therapeutic regimens should be made based upon positive PET FDG results after first-line treatment.

Radiation therapy can result in inflammatory changes that may be difficult to differentiate from recurrent tumor. Generally, radiation produces a diffuse, mildly elevated FDG accumulation within the tissues included in the radiation port [41]. This activity should decrease over time and scans are more reliable when obtained at least 12 weeks after completion of XRT [55]. Scans will be most reliable when a year or more has passed from the last radiation treatment [41]. A SUR of 2.5 still appears to be accurate in differentiating benign changes from tumor [41].

Recurrent Lung cancer: Differentiating Scar/Fibrosis from Recurrent Neoplasm:

Although no conclusive data support the use of any therapies in relapsed lung cancer, some patients with localized disease can be cured with aggressive therapy [2]. A major problem in the follow-up evaluation of patients treated for lung cancer is distinguishing post-treatment changes from recurrent tumor [48]. Presently, tumor recurrence is often not diagnosed until the disease has progressed to the point of marked enlargement of questionable abnormalities [41]. PET FDG imaging is very sensitive and highly accurate in distinguishing recurrent malignancy from scarring/fibrosis [13,41,48,49]. FDG PET imaging has been shown to have a sensitivity 98-100% for the differentiation of post-treatment change from tumor recurrence [48]. In a prospective evaluation of patients treated for NSCLC [48], FDG PET was able to correctly identify recurrent disease in all patients (sensitivity 100%) [48]. In this same group, CT imaging was non-contributory in 28% of cases at the time of FDG PET diagnosis [48].

Hicks et al also evaluated FDG PET in patients with suspected recurrent non-small cell lung cancer [2]. PET exams were positive in 41 of 42 patients with confirmed relapse (sensitivity 98%) [2]. The PET exam was negative in 14 of 17 patients with no evidence of relapse (specificity 82%), while CT was abnormal in 15 of these 17 patients (specificity 12%). Overall, CT was correct in identifying disease extent in 24% of patients, while PET was correct in 86% [2]. PET findings affected patient management in 63% of cases [2]. Patients with a positive PET exam had a poorer prognosis compared to patients with a negative study [2].

False positive exams can occur in association with infection and radiation pneumonitis [48]. Positive FDG PET exams due to radiation pneumonitis can be seen in up to 4% of patients receiving XRT therapy. The abnormality is usually diffuse, involves the lung and pleura, and conforms to the radiation port [2]. The duration of abnormality associated with radiation pneumonitis is variable and exams can remain positive (SUV greater than 2.5) for 6 to 15 months [48]. FDG PET findings associated with radiation pneumonitis should show decreased activity over time [48]. It is probably best not to perform FDG PET imaging earlier than 6 months after XRT in order to reduce the likelihood of false-positive PET results [48].

FDG-PET can also assist in identifying hypermetabolic foci within radiographic abnormalities to better direct tissue biopsy [45].

Screening for recurrence is another potential use of FDG PET imaging [48]. In a group of patients with NSCLC treated for cure, FDG PET screening was able to document the presence of recurrent tumor in all patients, whereas the finding was missed by CT in 44% [48]. If detected sooner, retreatment could be instituted in an effort to obtain prolonged disease control [48]. Further studies will be required to determine if improved detection of recurrent on FDG PET screening can result in improved patient survival [48].

Prognosis:

Among patients with resected non-small cell lung cancer (NSCLC) between 30% to 50% will develop recurrent tumor [6]. The TMN staging system provides some insight into overall patient prognosis, but fails to explain differences in survival among patients with similar staged lesions. Evaluation of tumor molecular biology has provided a greater understanding of why certain tumors are more likely to recur. In patients with NSCLC, measurements of tumor proliferation estimated by proliferating cell nuclear antigen (PCNA) and Ki-67 expression have prognostic value for recurrence and survival [6]. Increased PCNA and Ki-67 expression are associated with a poor outcome [6]. FDG uptake in NSCLC has also been correlated with tumor growth rate, aggressiveness, and proliferation capacity [6]. FDG uptake has been identified as an independent prognostic factor correlated with survival in patients with lung cancer- particularly early stage lesions [6]. Varying SUV cut-offs have been determined to select patients with the greatest risk for recurrence [6]. Essentially, the higher the SUV, the higher the proliferation capacity (aggressiveness) of the tumor and the worse the prognosis [6].