Positron Emission Tomography Cameras

A PET camera produces three-dimensional images that represent the distribution of radioactivity in the body. Any molecule that can be labeled with a positron-emitting radioisotope can be used to generate PET images (more than 400 PET tracers are listed in the NIH Molecular Imaging and Contrast Agent Database [MICAD], available at www.ncbi.nlm.nih.gov/books/NBK5330/).

The spatial resolution of older PET cameras was 6 mm or higher; for contemporary PET cameras, this is around 4 mm. Lesions with a diameter up to twice that resolution can be characterized with virtually no size-related (partial volume effect) underestimation of the tracer uptake, whereas for smaller lesions, the tracer uptake will gradually be underestimated as the size becomes smaller. In practice, lesions larger than 8 to 10 mm will be well characterized, whereas smaller ones, other than strongly FDG-avid lesions, cannot be accurately depicted.

The main difference between standard radionuclide imaging with gamma cameras and imaging with dedicated PET cameras is that the latter type of camera has a full ring of several thousands of scintillation detectors and does not need lead collimators—which absorb more than 99% of the emitted photons—to generate the image, resulting in higher sensitivity to radioactivity and higher spatial resolution.

Historically, PET cameras were “stand-alone” machines, either dedicated PET cameras or specially designed gamma cameras with which dual-head gamma camera coincidence imaging was performed. To overcome the lack of anatomic information of PET imaging, this type of camera has been replaced by hybrid systems in which a dedicated PET camera is combined with an anatomic tomograph—mostly with a computed tomography (CT) camera but sometimes a magnetic resonance imaging (MRI) camera. These fusion PET-CT cameras are considered the new standard (“stand-alone” PET cameras are not manufactured anymore), whereas PET-MRI is an emerging technology. The use of hybrid PET-CT cameras offers three main advantages: (1) attenuation correction (AC), which is needed to correct the image for the fact that some of the photons coming from radioactive decay are absorbed by the body, can be performed with the CT dataset, resulting in significant time reduction (approximately 10 minutes gained per patient); (2) increased accuracy of the exact position of the lesion and morphologic characterization of the underlying correlate, reducing equivocal findings; and (3) significantly increased confidence in reported findings. Typical scan times for modern PET-CT are in the 6- to 20-minute range for a skull-to-thigh image (i.e., whole-body scan). The data reported in this chapter derive from either dedicated PET or PET-CT applications.

The advent of PET-CT has resulted in two different strategies: so-called low-dose CT, which is used only for AC and localization, and “one-stop shopping” high-dose, contrast-enhanced diagnostic CT together with PET. It has been demonstrated that the use of oral or intravenous contrast agents does not induce clinically significant changes in the PET images. The combination of contrast-enhanced CT with PET changes tumor-node-metastasis (TNM) staging in 8% of patients and is nowadays mandatory for applications such as radiation therapy planning. The drawback is an increase in radiation dose, with low-dose techniques adding about 3 mSv to the approximately 8 mSv from the radiopharmaceutical, whereas contrast-enhanced CT adds some 10 to 20 mSv.

Metabolic Tracers

For cancer imaging, 2-¹⁸F-fluoro-2-deoxy-D-glucose (FDG), described in 1978, is by far the most commonly used metabolic tracer. The usefulness of this tracer relates to the increased cellular uptake of glucose (due to an increased expression of glucose transporter proteins) and a much higher rate of glycolysis in cancer cells. FDG, a glucose analogue in which the oxygen molecule in position 2 is replaced by a positron-emitting fluorine 18 atom, undergoes the same uptake as for glucose but is metabolically trapped and accumulated in the neoplastic cell after phosphorylation by hexokinase.

General Principles

If the aim of the FDG-PET study is just to stage the patient’s cancer, visual analysis of non-AC images (i.e., “hot spots” with higher-than-background activity not caused by physiologic processes are positive for tumor) probably is just as good as AC images, as has been pointed out by different prospective studies, both for the discrimination of nodules and for the evaluation of mediastinal involvement. Non-AC images should be examined to detect small lung lesions, because they are better visualized, owing to higher contrast, than are AC images.

The Normal FDG-PET Image

High physiologic FDG uptake occurs in brain, kidney, and urinary tract (urinary excretion) and can be present in the heart. Particularly in the brain, this interferes with lesion detection. A low degree of physiologic uptake of FDG has been noted in thoracic structures, including the lung, the heart, the aorta and large arteries, esophagus, thymus, trachea, thoracic muscles, bone marrow, and joints and soft tissues. This low background tracer activity builds the image contour.

False-Positive Results

FDG uptake is not tumor-specific and may be observed in all active tissues with high glucose metabolism, in particular, those in which inflammation is present. Therefore, a finding of clinically relevant FDG uptake, especially if isolated and decisive for patient management, requires confirmation. The differentiation between metastasis and a benign or inflammatory lesion, or even an unrelated second malignancy, should be made by means of other tests or tissue diagnosis.

The major causes of false-positive results (Box 8-2) in chest pathology are infectious, inflammatory, and granulomatous disorders. Iatrogenic procedures, such as thoracocentesis, placement of a chest tube, percutaneous needle biopsy, mediastinoscopy, thoracoscopy, and talc pleurodesis, also may give false-positive results.

False-Negative Results

False-negative results are less common and may be due to lesion-dependent or technical factors (see Box 8-2). A critical mass of metabolically active malignant cells is required for PET detection. Interpretation thus is a critical process with tumors exhibiting decreased FDG uptake such as small, very well-differentiated adenocarcinoma, bronchioloalveolar carcinoma, or carcinoid tumors. FDG-avid lesions smaller than 5 mm may be false-negative as a consequence of the limitations in spatial resolution and partial volume effect. In the lower lung fields, the detection limit may even go down to 10 mm, owing to additional respiratory motion. CT-based AC can cause artifacts in the event of misregistration between the CT and the PET data, which can lead to occultation of liver metastasis on the AC images.

Factors related to technique are paravenous FDG injection and high baseline glucose serum levels. Blood glucose levels should be checked, and it is advised to proceed only if the glucose level is within an acceptable range before tracer injection (typically 60 to approximately 180 mg/dL). Although diabetic patients often were excluded in the prospective studies, FDG uptake probably is not significantly influenced in these patients if the blood glucose levels are under reasonable control.

Tumor-Node-Metastasis Status

On the basis of extension of the primary tumor (T), spread to locoregional lymph nodes (N), and presence of distant metastasis (M), patients with lung cancer can be grouped according to different disease stages. Stage is the most important factor in prognosis and choice of treatment. Therefore, reliable noninvasive methods for accurate staging are very important. CT scan, endoscopic techniques, and surgical staging procedures are key staging tools, but addition of FDG-PET to these conventional methods has been shown to improve the staging process substantially, by distinguishing patients who are candidates for radical approaches such as surgical resection or intense multimodality treatments from those who are not.

For the T factor, the detailed images of modern multislice CT allow evaluation of the relationship of the tumor to the fissures (which may determine the type of resection), to mediastinal structures, or to the pleura and chest wall. The integrated FDG-PET–CT images (Figure 8-2) may allow more precise evaluation of chest wall and mediastinal infiltration or correct differentiation between tumor and peritumoral inflammation or atelectasis.

Figure 8-2 ¹⁸F-fluorodeoxyglucose positron emission tomography–computed tomography (FDG-PET–CT) aids in the determination of T stage. Patient with adenocarcinoma of the left lower lobe. A, CT images show densification of the left lower lobe due to a solid lesion with accompanying retro-obstructive atelectasis, without clear delineation between tumor and benign parenchyma. B, Coronal image of the FDG-PET–CT shows a large rounded hypermetabolic lesion in the left lower lobe with smaller foci of intense activity in the mediastinum medial and cranial to the tumor, corresponding to lymph nodes. C, Transverse PET image showing a nicely delineated rounded tumor mass, allowing appropriate size determination (47 mm, or T2a). D, Corresponding low-dose CT image. E, Corresponding fusion image. F, Coronal image showing limited size of tumor compared with the atelectatic region. G, Consecutive fusion images allowing determination of involved and spared tissue, which are of high value in radiation therapy planning.

FDG-PET–CT has been used to assess pleural disease with variable results, because small pleural deposits can be missed on PET-CT, owing to their low tumor load or partial volume effect, whereas false-positive results have been documented in patients with inflammatory pleural lesions. If pleural abnormalities determine the chance for radical treatment, often histopathologic verification with cytologic analysis or thoracoscopic biopsy is needed.

For the N factor, the addition of FDG-PET to CT results in more accurate lymph node staging than CT alone, with an overall sensitivity of 80% to 90% and a specificity of 85% to 95% in metaanalyses.

On the one hand, the absence of mediastinal lymph node disease on FDG-PET–CT has a high NPV, so that invasive lymph node staging tests can be omitted, and the patient can proceed to straightforward surgical resection. Restrictions on this NPV apply in case of insufficient FDG uptake in the primary tumor, central location of the tumor, or presence of important hilar nodal disease that may obscure coexisting N2 disease on PET images.

On the other hand, the combination of FDG-PET and CT illustrates the location of suspect lymph nodes, thereby helping to direct tissue sampling procedures such as endobronchial ultrasound–guided transbronchial needle aspiration or cervical mediastinoscopy. Because false-positive results are possible with lymph node imaging (under the conditions listed in Box 8-2), histopathologic proof of lymph node involvement should be sought in most patients with imaging-positive mediastinal nodes on FDG-PET, except those with obvious bulky nodes on imaging (Figure 8-3).

Figure 8-3 ¹⁸F-fluorodeoxyglucose positron emission tomography–computed tomography (FDG-PET–CT) in the determination of nodal (N) stage. Patient with a squamous cell carcinoma of the left lower lobe. A, Coronal image shows hypermetabolic tumor in left lower lobe with a hypometabolic area at the cranial part indicating necrosis, hypermetabolic bulky mediastinal nodes, and focal increased uptake in the left supraclavicular region. B, CT image of primary tumor. C, Fusion image of primary tumor. D, CT image of bulky enlarged subaortic nodes and calcified rib lesion. E, Fusion image demonstrates hypermetabolic nodes and absence of metabolic activity in the rib. F, Nonenlarged supraclavicular lymph node behind the jugular vein. G, Focal intense FDG uptake in this node (SUV_max = 12.2) is indicative of N3 involvement. The benign nature of the rib lesion was confirmed through biopsy, and the malignant nature of the supraclavicular node, by fine needle aspiration. SUV, standardized uptake value.

For the M factor, FDG-PET added to CT is almost uniformly superior to CT alone, except in brain imaging, for which sensitivity is unacceptably low owing to the high glucose uptake in normal surrounding brain tissue. CT and, even better, MRI remain the methods of choice for brain imaging.

For bone metastases, FDG-PET is more accurate than technetium 99m–tagged medronate disodium (formerly methylene diphosphonate) (^99mTc-MDP) bone scan: Sensitivity is at least as good (90% to 95%), and specificity is far better (95%, versus 60% for bone scan). Limitations are the restricted area of imaging for PET (only from the head to just below the pelvis) and possibility of a false-negative result with osteoblastic lesions (rare in lung cancer). For adrenal gland metastases, FDG-PET has a high sensitivity for detection of such disease, so an equivocal-appearing lesion on CT without FDG uptake usually will not be metastatic. FDG-PET also can be of help for investigation of hepatic lesions that remain indeterminate by conventional studies. PET also may reveal metastases in sites that escape attention on conventional staging (e.g., soft tissue lesions, retroperitoneal lymph nodes, barely palpable supraclavicular nodes, painless bone lesions). Exclusion of malignancy requires caution in case of smaller lesions (less than 1 cm in diameter) (see Box 8-2). In this respect, the small contralateral lung nodule(s)—a common finding in the era of spiral multislice CT—often still requires invasive sampling, such as by thoracoscopy.

Influence on Treatment Choices

FDG-PET has a significantly complementary role to that of CT for two reasons: First, PET may detect unexpected lymph node or distant organ metastatic spread (Figure 8-4). After a negative result on conventional staging, unknown metastases are found on PET-CT in 5% to 20% of the patients, in increasing numbers from clinical stage I to III tumors. Second, FDG-PET is able to determine the nature of some equivocal lesions on conventional imaging. Consequently, PET induces a change of stage in 27% to 62% of patients with non–small cell lung carcinoma (NSCLC), in general more upstaging than downstaging, related mainly to the detection of unexpected distant lesions (Table 8-1). Such findings lead to a change in management in 25% up to 52% of the patients. Documented changes have involved both treatment intent (curative versus palliative) and choice of treatment modalities (chemotherapy versus radiotherapy, radical radiotherapy versus surgery).

Figure 8-4 ¹⁸F-fluorodeoxyglucose positron emission tomography–computed tomography (FDG-PET–CT) may detect unsuspected metastases. Patient with large cell carcinoma of the right upper lobe assessed with FDG-PET–CT after concurrent chemotherapy and radiotherapy. A, Coronal image shows a persistent, intensely hypermetabolic mass in the right upper lung. Pronounced linear FDG uptake in the esophagus corresponded to treatment-induced inflammation. Two additional foci project on the midline of the epigastrium (arrows). B, Transverse PET image of the cranial epigastric lesion. C, Fusion image shows an intra- or juxtapancreatic location, highly suspicious for metastasis. D, Corresponding CT image shows no significant lesion. E, Hypointense focal intrapancreatic lesion on a subsequent contrast-enhanced T1-weighted MRI study, compatible with metastasis (arrow). F-I, Similar findings in the caudal lesion. Echoendoscopy-directed fine needle aspiration confirmed metastasis. MRI, magnetic resonance imaging.

No problem arises in interpretation when whole-body FDG-PET shows multisite metastases, but presence and character of an isolated suspect lesion that determines radical treatment intent should always be verified by other tests or tissue sampling, because of the ever-present risk of false-positive lesions (see Box 8-2) or a second primary tumor. In one large retrospective series, solitary extrathoracic lesions were documented in approximately 20% of the patients. About half of these were metastatic; the other half were not related to lung cancer, as either inflammatory or other benign lesions, or were second primary tumors (Figure 8-5).

Figure 8-5 ¹⁸F-fluorodeoxyglucose positron emission tomography–computed tomography (FDG-PET–CT) detects synchronous tumors, as in this patient with adenocarcinoma in the right upper lobe. A, Coronal image shows focal intense uptake in the right upper lobe as well as in the left breast. Moderately increased uptake is apparent in a right paratracheal lymph node. Note multifocal, moderately increased uptake in the right ribs (secondary to inflammation due to recent fractures) and next to the right trochanter femoris (enthesopathy). B, Transverse PET image shows the intensely FDG-avid lung lesion, compatible with the known malignancy. C, CT image of lung lesion. D, Fusion image. E, Transverse PET image of the intensely FDG-avid focal lesion in the left breast, highly suspicious for malignancy. F, CT shows an enhancing solid lesion at the site of FDG uptake. G, Fusion image of breast lesion that proved to be an invasive ductal adenocarcinoma. H, Coronal image after induction chemotherapy for the pulmonary adenocarcinoma and antihormonal therapy for the breast lesion. Note the strong decrease in uptake in both primary tumors, indicating a partial metabolic response. The intense focus in the left supraclavicular region is an artifact due to tracer retention in the patient’s Portacath.

Before the advent of integrated FDG-PET–CT, it remained unclear if PET alone could replace conventional imaging. One randomized study compared staging with upfront FDG-PET alone (i.e., directly after first presentation) versus guideline-based CT imaging in 465 patients. Patients with FDG-avid, noncentrally located tumors without signs of mediastinal or distant spread on PET proceeded directly to surgical resection. Frequency of noninvasive tests to reach a clinical TNM diagnosis was about the same in both treatment groups, but that of invasive tests (i.e., mainly mediastinoscopy) was significantly reduced with PET. With the contemporary use of integrated FDG-PET–CT, this question has become an academic one.

The effect of adding FDG-PET or FDG-PET–CT was investigated in several randomized controlled trials (Table 8-2). Two earlier trials that looked at the addition of PET alone reported seemingly contradictory results. Clear differences in trial design probably accounted for this finding. In the Dutch trial, the end point “futile thoracotomy” was clearly defined (benign disease, explorative thoracotomy, pathologic stage IIIA-N2/IIIB, or postoperative relapse or death within 12 months), whereas in the Australian study, there were no benign lesions, surgery was considered to be of use in some patients with stage IIIA-N2 disease, and no strict follow-up terms were predefined. The Australian trial focused on patients with clinical stage I and II disease only, where less additional benefit of PET was demonstrated in previous nonrandomized accuracy studies. The later trials used FDG-PET–CT imaging in addition to standard workup. The study by Fischer and colleagues largely reproduced the Dutch experience, with a significant reduction in futile thoracotomies. However, the number of nonfutile thoracotomies and the overall survival rate were similar in both treatment groups. Finally, the study reported by Maziak and associates mainly looked at correct upstaging and met this primary end point. Even in this study, which was the largest, no significant survival difference at 3 years was noted. This apparent lack of improvement was attributed by the investigators to the fact that larger patient numbers may be needed, and to the fact that some patients did not have their planned surgery.

The overall evidence thus points at significantly more accurate TNM staging with PET-CT compared with conventional imaging alone. This improved accuracy leads to more appropriate treatment policies, such as the avoidance of futile surgery in more patients than when only conventional imaging is used. In nonrandomized settings, this also has been reported to lead to other treatment adaptations—for example, a change between radical and palliative intent of radiotherapy, or adaptations in the radiation fields. Although all of this evidence certainly leads to true benefits such as stage migration and better patient management, it remains unknown if survival of individual patients has improved since the introduction of FDG-PET.

Survival

Indeed, several FDG-PET staging studies clearly demonstrated stage migration. The possible effects of stage migration may account in part for improvements in survival of patients treated for both early- and advanced-stage disease, widely referred to as the “Will Rogers phenomenon.” Examples are the prospective multicenter trials on the use of FDG-PET in patients with stage III NSCLC, in which upstaging was confirmed in about 25% to 30% of the patients, and in which overall survival was significantly longer (P = .006) in patients whose disease was staged by PET than in those who did not undergo PET. The aforementioned randomized controlled trials were underpowered to evaluate the potential real survival benefit brought by PET.

FDG-PET also has been shown to predict the prognosis for patients with NSCLC. A recent systematic review and meta-analysis including retrospective studies demonstrated that the SUV, a semiquantitative measure of FDG uptake, for the primary tumor at diagnosis may predict outcome in NSCLC, especially at earlier stages. These studies almost consistently found a better overall survival among patients with a metabolic activity lower than the threshold SUV value, calculated from either the most discriminative log-rank SUV value or the median SUV. However, although SUV may be a way to assess prognosis, no true cutoff point has been recognized as suitable for broad clinical use. Rather, a continuous SUV spectrum of a gradually worsening prognosis might be a more realistic concept. Baseline SUV, incorporated as a continuous variable in a Cox proportional hazards model, showed a 7% increase in hazard of death after a one-unit increase in SUV in patients with resected stage I to III NSCLC and a 6% increase in hazard of death after a one-unit increase in SUV in patients with inoperable NSCLC treated with radiotherapy.

Health Economics

Today’s respiratory oncologists aim for the best-quality health care for the patient, but in acknowledgment of the need for financial prudence. The major cost of modern oncology practice, however, does not lie in the baseline diagnostic process but is attributable to the delivery of expensive treatments and the morbidity related to possible side effects. Therefore, application of economic modeling in the use of FDG-PET has to be based on both diagnostic and therapeutic aspects of health care expenditure within the day-to-day clinical setting.

In a recent overview of all economic evaluations on FDG-PET in oncology performed between 2005 and 2010, the strongest evidence for cost-effective use of PET alone was for the staging of NSCLC. Studies suggested that PET for staging of NSCLC may benefit patients in terms of a (slight) increase in life expectancy and may benefit the health care system in terms of cost savings resulting from the number of invasive procedures avoided. Since the introduction of PET-CT technology into clinical medicine in 2001, only a few additional studies in respiratory oncology have evaluated the cost-effectiveness of this integrated scanning method. Taking into account the superior accuracy of FDG-PET–CT over that of PET alone in lung cancer staging, the health economic impact in terms of cost-effectiveness probably can be extended to PET-CT. Furthermore, FDG-PET is cost-effective for characterizing and diagnosing solitary pulmonary nodules and constitutes the most cost-effective strategy for assessment of nodules of low to moderate pretest probability of malignancy on CT.

Other Indications for Positron Emission Tomography

Mesothelioma

Integrated FDG-PET–CT imaging is playing an increasing role in the assessment of suspected or known malignant pleural mesothelioma (MPM) (Figure 8-6). PET-CT is useful in the correct differentiation of malignant (mainly MPM) from benign pleural diseases in asbestos-related CT findings, with an overall accuracy greater than 90% and a high NPV of more than 90%. PET-CT is significantly more accurate in baseline TNM staging in patients who are appropriate candidates for multimodality therapy based on spiral CT findings alone. Although PET-CT does not provide additional information about the primary tumor beyond that supplied by CT alone, it identifies a higher number of metastatic mediastinal lymph nodes and/or unknown distant metastatic disease in up to two thirds of patients, with a significant clinical impact on treatment planning. Early evidence also suggests that PET-CT may have a role in evaluating response to therapy in MPM. This is an interesting possibility, because the assessment of response in patients with MPM according to standard response criteria on CT is far from simple. More work to define response criteria for MPM on FDG-PET is needed. Furthermore, a prospective study in patients with nonsarcomatoid malignant pleural mesothelioma observed that baseline total glycolytic volume on PET was more predictive of survival than CT-assessed TNM stage on multivariate analysis. These observations of prognostic capability still require prospective validation.

Figure 8-6 ¹⁸F-fluorodeoxyglucose positron emission tomography–computed tomography (FDG-PET–CT) for evaluation of a patient with right-sided epithelioid pleural mesothelioma. Pleuroscopy with talc pleurodesis was performed before the PET scan. A, Coronal image showing diffuse and irregular area of intense FDG uptake in the right hemithorax, mainly in the apical part but with also demonstrating multiple foci in the costodiaphragmatic sinus. B, Coronal PET image showing hypermetabolic activity in the lateral and medial right pleura. C, Fusion images show areas of pleural effusion without increased FDG uptake.