DEFINITIONS
- ▪
SUV : Refers to the ratio of the concentration of radiopharmaceutical (usually FDG) in a volume of tissue in microcuries of injected agent per volume to concentration in the body if uniformly distributed (determined by a standard body phantom). The SUV has no units. An SUV of 1.0 is achieved in any tissue volume when the count rate is equal to the count rate of the uniformly distributed activity in the body phantom. The results are usually normalized to body weight.
- ▪
Positron: Positive electron
- ▪
Coincident event: Annihilation of matter when positron combines with a negative electron resulting in two 511 keV photons originating at 180 degrees from each other. The two photons are detected by the PET scanner as a coincident event if they strike the detector within a narrow time window (7–10 nanoseconds) with a 511 keV energy
- ▪
FDG: Glucose analog that is transported across the cell membrane similar to glucose by transporters such as GLUT-1. FDG is phosphorylated by hexokinase similar to glucose and cannot leave the cell ( Fig. 2-1 ). FDG is not able to be metabolized in the glycolytic pathway like glucose. In highly metabolic tumors with a high requirement for glucose, an FDG concentration far greater than normal background is usually achieved.
Figure 2-1
A schematic of glucose and fluorodeoxyglucose (FDG) uptake by a tumor cell. Note the GLUT-1 transporter recognizing both for cell entry while both are phosphorylated by hexokinase. Glucose is able to be metabolized, whereas FDG cannot enter the glycolytic pathway and is trapped within the cell, resulting in a high concentration of the radiotracer and is easily detected using the modern PET scanners.
IMAGING CONSIDERATIONS IN THE SOLITARY PULMONARY NODULE
- ▪
FDG PET
- ○
Malignant solitary pulmonary nodules (SPNs) tend to have positive PET when they are greater than 10 mm
- ▪
Sensitivity for nodules greater than 10 mm is 95% to 98%.
- ▪
- ○
Spiculated nodules on computed tomography (CT) with strongly positive PET (SUV > 2.5) usually malignant
- ▪
Specificity for malignancy increases with increasing SUV ( Fig. 2-2 ).
Figure 2-2
A 68-year-old woman, with a history of smoking, living in endemic area for histoplasmosis. CT ( A ) demonstrates scar-like abnormality in the left upper lobe. FDG-PET ( B ) demonstrates intense activity (SUV = 6.8) suggesting malignancy. Pathologic diagnosis is squamous cell cancer of the lung.
- ▪
- ○
- ▪
Sensitivity
- ○
Small (<10mm) malignant pulmonary nodules seen on CT may be negative on FDG-PET or have a reduced SUV due to a partial volume effect.
- ▪
Partial volume effect will lower the perceived activity of small objects (size below system resolution).
- ▪
Lesions as small as 4 to 5 mm are detectable with FDG-PET, but accurate SUVs are not available unless corrections are made using phantom generated solutions ( Fig. 2-3 )
Figure 2-3
Synchronous primaries (1.2 cm and 0.9 cm). Note the small PET-positive nodule anterior to the larger nodule in the right upper lung. SUVs in small nodules (<10 mm) may be artificially reduced due to partial volume effects.
- ▪
- ○
Sensitivity and negative predictive values for indolent malignancies such as carcinoid tumors, and bronchoalveolar and well-differentiated adenocarcinomas are less than tumors with high metabolic requirements ( Fig. 2-4 ).
Figure 2-4
A, Coronal FDG-PET views of patient with a ground-glass opacity (3.5 cm) in the right upper lung and a 2.0-cm nodule in the left upper lung. Pathology is hamartoma on the left, with inflammatory cells and bronchoalveolar cancer of the right upper lung. B, CT with PET axial slices in same patient as Figure 2-4A .
- ○
Sensitivity is reduced in “stunned” tumors during or following chemotherapy.
- ○
Tumor uptake is strongly influenced by glucose and insulin levels. High glucose levels may significantly reduce tumor SUV ( Fig. 2-5 ).
Figure 2-5
Importance of proper patient preparation before FDG-PET scan. Patient with elevated glucose demonstrates low-level activity in right lower lung primary adenocarcinoma. Patient properly prepared (no carbohydrate for a minimum of 6 hours before injection of FDG) demonstrates increased intensity in primary with additional foci in both right and left lung.
- ○
- ▪
Specificity
- ○
Specificity and positive predictive value of 50% to 80% are influenced by patient selection, regional influence with endemic fungal disease, exposure to tuberculosis, environmental exposures, medication, as well as criteria used to define a positive study.
- ○
- ▪
Accuracy
- ○
FDG-PET accuracy is greater than that of CT for characterizing SPN.
- ○
- ▪
Prognosis
- ○
SUV values for an SPN of 2.5 have been suggested as a dependable marker for malignancy ; however, many benign lesions will have SUVs at this level, and sensitivity and specificity values at this level may be unreliable, especially in regions with high prevalence of tuberculosis or fungal disease.
- ○
SUV values of 5 to 10 for malignant SPNs are highly associated with a poor prognosis.
- ▪
Reports indicate that FDG-PET is more accurate than pathology in predicting survival or recurrence.
- ▪
- ○
Staging with FDG-PET is effective in offering prognostic information. However, even with a PET-negative study other than the pulmonary nodule (Stage I PET), an SUV greater than 5 is associated with poor patient outlook.
- ○
- ▪
Problem areas in SPN evaluation with FDG-PET
- ○
False-positive studies for FDG-PET are seen in many conditions.
- ▪
Infectious granulomas including fungal, tuberculosis
- ▪
Other infections; aspergillosis, bacterial, viral
- ▪
Noninfectious granulomas; sarcoid, rheumatoid nodules, Wegener’s
- ▪
Benign neoplasms; hamartoma, fibroma
- ▪
Inhalation disease; silicosis, lipoid pneumonia
- ▪
Other pathology; pulmonary infarction, sequestration, inflammatory alveolitis
- ▪
- ○
- ▪
Dual-time point imaging
- ○
Increasing SUV over time (60 versus 180 minutes) suggests malignancy.
- ○
Decreasing SUV over time suggests benign nodule.
- ○
Key Facts
- ▪
Probability of malignancy for SPN increases with increasing FDG-PET activity
- ▪
Any FDG activity in a spiculated nodule on CT suspicious for malignancy especially if nodule increases in size on serial x-ray study or CT within 90 days
- ▪
High sensitivity for 15 mm or larger pulmonary nodules due to non–small cell lung cancer (NSCLC) (95–98%)
- ▪
Specificity of FDG PET for NSCLC pulmonary nodule 50-75% dependent on multiple factors
- ▪
Indolent pulmonary malignancies have low SUVs and may be visually PET negative (bronchoalveolar cancer, carcinoid, well-differentiated adenocarcinoma)
- ▪
Correct patient preparation (low-carbohydrate diet for a minimum of 6 hours before injection of FDG and no exercise for 24 hours recommended for optimum results).
- ▪
Dual-time point imaging may be helpful in differentiating a benign from a malignant nodule.
FDG-PET/CT IN STAGING OF NON–SMALL CELL LUNG CANCER
- ▪
Accurate staging of subjects with NSCLC is required for optimum patient management.
- ▪
CT staging for NSCLC is inaccurate due to size criteria of 10 mm for positive node.
- ▪
FDG-PET, especially with CT fusion, has emerged as the most accurate noninvasive approach to staging.
- ▪
FDG-PET has been shown to be a more accurate indicator of nodal or distant metastatic disease than imaging with CT or MRI alone.
- ▪
FDG-PET sensitivity for detection of nodal spread of NSCLC has been reported to range from 75% to 95%. Variability varies due to instrumentation, patient selection bias, and surgical and pathologic sampling ( Fig. 2-6 ).
Figure 2-6
Coronal FDG-PET views demonstrating right upper lung focal abnormality with right lower paratracheal focus as well. Pathology is adenocarcinoma with mediastinal metastasis (Stage 3A).
- ▪
Probability of nodal malignancy is greatest when positive nodes are larger than 10 mm ( Fig. 2-7 ), followed by FDG-positive nodes smaller than 10 mm, followed by FDG-negative nodes larger than 10 mm, and least probable when FDG-PET is negative in nodes smaller than 10 mm.
Figure 2-7
Maximum Intensity Projection (MIP) FDG-PET image in a patient with right upper lung adenocarcinoma and extensive mediastinal metastasis. Primary tumor SUV = 12.4. Higher SUV values of primary generally correlate with poor prognosis and increased frequency of metastasis.
- ▪
FDG-PET is more accurate than CT in nodal staging; however, use of combined PET/CT systems or software fusion is more accurate than PET or CT alone ( Fig. 2-8 ).
Figure 2-8
A, Patient with 3.5-cm spiculated right upper lung lesion on CT scan, and borderline precarinal lymph node, otherwise normal. FDG-PET demonstrated intense activity in the lesion (SUV = 9.2) and also detected a small mediastinal node. Staging increased from 2A to 3A. Pathology is adenocarcinoma with positive precarinal lymph node. B, CT scan demonstrating borderline precarinal lymph node with definite FDG activity on FDG-PET.
- ▪
Specificity may be reduced in regions of endemic tuberculosis (TB) and fungal disease.
- ▪
PET correctly alters staging 20% to 30% of the time and changes patient management 14% to 37% of the time ( Fig. 2-9 ).
Figure 2-9
A, Large right lower lung nodule 4.2 × 3.5 cm. SUV is 10.4. Note metastasis to the left acetabulum. Pathology is squamous cell cancer of the lung. B, Same patient in Figure 2-8A . Note metastatic focus adjacent to spine and posterior mediastinum.
- ▪
PET staging is a better predictor of time to death than clinical or CT staging.
Surgical Resectability
- ▪
Accurate staging important in determining resectability for cure.
- ▪
T 3 lesions (extension into chest wall, diaphragm, mediastinal pleura, pericardium or main bronchus) may have curative surgical potential.
- ▪
T 4 lesions (invasion of mediastinum, great vessels, heart, trachea, esophagus, vertebra, or other vital structures) generally not amenable to surgery.
- ▪
PET specificity and positive predictive values are unreliable for mediastinal evaluation with unacceptable false-positive results ( Fig. 2-10 ). Tissue confirmation is required in most cases of positive mediastinal PET results. Usually, this is done with mediastinoscopy when possible.
- ○
PET will help to determine the best biopsy approach and also may determine if stage IV disease is present.
- ○
Central primary tumor may obscure mediastinal nodal activity, especially with high SUV values (mediastinoscopy suggested).
Figure 2-10
MIP image with solitary left upper lung pulmonary nodule. Multiple small foci in bilateral hilum and mediastinum. Pathology for nodule is adenocarcinoma. Pathology for mediastinal nodes is noncaseating granulomas. FDG-PET detects small lymph nodes with granuloma or other inflammatory/infectious processes and is therefore not specific for tumor. The distribution pattern in this patient is typical for granuloma.
- ○
Detection of Extrathoracic Metastases
- ▪
FDG-PET and PET/CT whole body scan is effective in detecting stage IV disease ( Fig. 2-11 ).
Figure 2-11
Patient with metastatic squamous cell cancer (Stage 4). Note multiple metastatic sites including mediastinum, axilla, and multiple osseous sites. Patient was asymptomatic, with no history of bone pain. FDG-PET has been shown to be an excellent test for staging of distant metastasis, and is superior to x-ray and CT imaging.
- ▪
The frequency of extrathoracic metastases is 10% to 15%.
- ▪
Lytic or trabecular metastases best detected with FDG-PET compared with conventional technetium-99m methylene diphosphonate (Tc99m MDP) bone scans.
- ▪
Osteoplastic bone metastases are best detected with conventional bone scans.
- ▪
Response to therapy is best assessed with FDG-PET as false-positive increases often seen with conventional bone scans due to healing effect (FLARE).
- ▪
Nonosseous metastases (adrenal and liver most common sites).
- ▪
Brain metastasis may be detected; however, due to high concentration of FDG, MRI is the test of choice
Restaging Following Therapy (Chemotherapy)
- ▪
Traditional response assessment by size changes with radiography, CT, or MRI.
- ▪
New response assessment uses morphology and function (tumor viability).
- ▪
Reduction in metabolic activity after one to three cycles of chemotherapy correlates with clinical response in most reports ( Fig. 2-12 ).
Figure 2-12
FDG-MIP images in a patient with widespread lung cancer pre- and postchemotherapy. A, Baseline prechemotherapy. B, After three cycles of chemotherapy.
- ▪
Management changes due to results of FDG-PET are as high as two thirds of cases.
Radiation Therapy Planning
- ▪
FDG-PET/CT allows contouring of target volumes using both anatomic and metabolic borders. Regions that are not detected with planning CT often are detected with PET allowing inclusion in radiation field ( Fig. 2-13 ).
Figure 2-13
FDG-PET fused with radiation therapy planning CT. Red areas on axial image indicate the target volume determined by contouring of CT. Green areas represent FDG-PET distribution. Orange regions represent overlap of CT contour and FDG-PET distribution. Note the tumor sites not included in the original CT contour map. FDG fusion mapping is frequently included in determining target volumes for radiation therapy.
- ▪
Changes in target volumes may increase or decrease depending on results of the FDG-PET scan
- ▪
Response to radiation therapy can be successfully monitored with FDG-PET.
- ○
Post-therapy inflammation, however, may persist for months, resulting in an inaccurate interpretation.
- ○
MESOTHELIOMA
- ▪
CT conventional imaging is the main imaging technique for assessment of malignant pleural mesothelioma.
- ▪
Limitations of CT
- ○
N staging is limited: enlarged nodes (>1 cm) are often benign. Small nodes (<1 cm) may harbor metastases.
- ○
Pleural thickening may be benign scarring.
- ○
Underestimates extent of chest wall involvement and peritoneal involvement.
- ○
- ▪
PET and PET/CT identifies malignant sites ( Fig. 2-14 ).
- ○
SUV higher in malignant sites than in most benign processes.
- ○
Superior to CT alone in nodal staging.
- ○
Identifies extrathoracic metastases.
- ○
Directs biopsy to most appropriate site.
- ○
High SUV is associated with a poor prognosis.
Figure 2-14
FDG-PET/CT in a patient with malignant mesothelioma (epithelioid type) involving the pleura of the left lower lung (SUV = 9.1). Fusion imaging is helpful in defining the extent and location of malignant mesothelioma. Pleurectomy via thoracoscopy demonstrated a malignant pleural mesothelioma. The pericardium was not involved.
- ○
VENTILATION/PERFUSION LUNG SCANS
Ventilation Scans
- ▪
Ventilation studies with gaseous radiopharmaceuticals
- ○
Xenon 133
- ○
Xenon 127
- ○
Krypton 81m
- ○
- ▪
Xenon 133 advantages
- ○
May be used pre or postperfusion.
- ○
Readily available commercially.
- ○
Inexpensive.
- ○
Wash-in, washout information.
- ○
Useful to establish postpneumonectomy stump leak.
- ○
- ▪
Xenon 133 disadvantages
- ○
Limited views
- ○
May be difficult for seriously ill or elderly patients to cooperate for study
- ○
Leakage from face mask may cause local contamination
- ○
- ▪
Xenon 127 advantages
- ○
Photopeak energy (keV) ideal for postperfusion ventilation studies
- ○
- ▪
Xenon 127 disadvantages
- ○
Not readily available
- ○
Expensive
- ○
Other as per Xenon 133
- ○
- ▪
Krypton 81m advantages
- ○
Short half life (13 sec) approximates alveolar distribution
- ○
- ▪
Krypton 81m disadvantages
- ○
Production generator (Rb81) required
- ○
Expensive
- ○
Not readily available
- ○
Not adequate to evaluate obstructive airway disease
- ○
- ▪
Ventilation studies with submicronic particles
- ○
Technetium-99m diethylenetriamine-pentacetic acid (Tc 99m DTPA)
- ○
Tc 99m sulfur colloid
- ○
- ▪
Tc 99m DTPA advantages
- ○
Multiple views are easy to obtain.
- ○
Study is inexpensive.
- ○
Patient cooperation is usually satisfactory.
- ○
Nebulizer kit and radiopharmaceutical are readily available.
- ○
Obstructive airway disease is really diagnosed.
- ○
- ▪
Tc 99m DTPA disadvantages
- ○
Difficult to use when performing a V/Q study with quantitation of perfusion unless perfusion is done before ventilation.
- ○
- ▪
Tc 99m sulfur colloid advantages
- ○
Similar to Tc 99m DTPA
- ○
- ▪
Tc 99m sulfur colloid disadvantages
- ○
Higher patient dose than Tc99m DPTA (slower clearance)
- ○
Slightly higher cost than Tc99m DTPA
- ○
Perfusion Lung Scan
- ▪
Radiopharmaceutical Tc 99m macroaggregated albumen (MAA) particles (30–60 μ size)
Perfusion Lung Scan: Clinical Applications
- ▪
Evaluation of pulmonary embolus (PE)
- ▪
Preoperative assessment before lung resection for tumor or parenchymal disease
- ▪
Prediction of postresection pulmonary function
- ▪
Evaluation of lung transplant candidate
- ▪
Evaluation of post-transplant pulmonary function in suspected rejection
- ▪
Evaluate right-to-left cardiac shunts
- ▪
Evaluate congenital heart disease
V/Q Evaluation of Pulmonary Embolus
- ▪
Prospective investigation of pulmonary embolism diagnosis (PIOPED): widely used in original or modified form to evaluate V/Q studies
- ○
Probability of PE assigned based on criteria established by V/Q scan correlation with angiography
- ○
Interpretation of V/Q Scan for PE Diagnosis
- ▪
Normal: No PE
- ▪
Near-normal: less than 5% probability
- ▪
Low probability (5–19%)
- ▪
Intermediate (20–79%)
- ▪
High probability (≥80%)
- ▪
Normal
- ○
Uniform activity throughout both lungs ( Fig. 2-15 )
Figure 2-15
Normal ventilation/perfusion lung scan. Ventilation was performed using a submicronic mist of Tc-99m DTPA (1 millicurie). Note the homogeneous appearance throughout both lungs. A perfusion lung scan is then performed with 4 millicuries of Tc-99m macroaggregated albumin particles (30–50 μ). Note the homogeneous appearance.
- ○
- ▪
Low probability
- ○
Small scattered defects
- ○
V/Q multiple matched findings with no infiltrates on chest x-ray study ( Fig. 2-16 )
Figure 2-16
Low-probability V/Q scan in a patient with extreme shortness of breath. Note the severe nonhomogeneous pattern on ventilation consistent with obstructive pulmonary disease and probable parenchymal disruption. Perfusion is mainly matching but overall appears more uniform. There are no mismatches. This finding is representative of patients with a pulmonary parenchymal process with an obstructive component.
- ○
Single matched moderate or large defect may be low or intermediate
- ○
Perfusion defects smaller than chest x-ray study findings
- ○
- ▪
Intermediate probability
- ○
Neither high or low probability
- ○
Mismatched defects less than 2 segmental equivalents
- ○
Matched V/Q and chest x-ray study
- ○
Knowledgeable interpreter should stratify within intermediate category (low to high intermediate)
- ○
- ▪
High probability
- ○
Two or more mismatched segments or segmental equivalents ( Fig. 2-17 )
Figure 2-17
High-probability V/Q scan demonstrating normal ventilation and multiple segmental and large subsegmental perfusion abnormalities. Mismatched segments or large subsegments greater than 2 segmental equivalents are considered a high probability for pulmonary emboli (greater than 80%).
- ○
- ▪
False positive V/Q for PE
- ○
Vasculitis
- ○
Central tumor involving pulmonary artery segments directly or nodal enlargement with vascular compression
- ○
Fibrosing mediastinitis
- ○
Chronic or prior PE with residual fibrinous webs
- ○
Primary pulmonary hypertension
- ○
Congenital defects involving pulmonary artery
- ○
Preoperative Assessment Before Resection with Perfusion Lung Scan
- ▪
Prediction of residual lung function postresection (calculation of percentage perfusion remaining approximates residual lung function).
- ▪
Identifies regional abnormalities where resection may give favorable results (upper lobe hypoperfusion identifies patients who are likely to improve with upper lobe resection allowing expansion).
- ▪
Quantitation techniques variable
- ○
Three-zone technique is not anatomic and unreliable.
- ○
Identify and quantitate upper lobe, lower lobe, middle lobe, and lingula.
- ○
Give percentage perfusion for each lung (normal 55–60% right, 40–45% left)
- ○
Within each lung, calculate percentage perfusion for each lung (lateral and oblique views helpful) ( Fig. 2-18A to D )
Figure 2-18
A, Perfusion lung scan with 4 mCi Tc-99m MAA particles in a 66-year-old patient with chronic obstructive pulmonary disease and emphysema for possible lung reduction surgery. B, Quantitative perfusion analysis in same patient; note the nonhomogeneous pattern throughout the lungs, with greater activity in the upper lung zones than in the mid and lower regions. Patient is not a candidate for a lung reduction surgery. Perfusion lung scans in a 61-year-old patient with emphysema. Note the bilateral reduction of upper lobe perfusion. D, Quantitative perfusion analysis of same patient demonstrates severe reduction in upper lobe perfusion bilaterally (right upper lobe, 12% of total right lung perfusion; left upper lobe, 11% of total left lung perfusion).
- ○
Right-to-Left Intracardiac Shunt Evaluation with Perfusion Lung Scan
- ▪
Pulmonary to systemic shunt usually interatrial or interventricular defect.
- ▪
Normally more than 95% of injected macroaggregated particles are trapped in lung capillaries.
- ▪
If right to left shunt is present, particles are trapped in extrapulmonary sites in proportion to blood flow, mainly kidneys and brain ( Fig. 2-19A and B )
