In PET, a biologically active molecule is labeled with a radionuclide and introduced into the body. Most of the positron-emitting radionuclides are produced in a cyclotron and have very short half lives and large positron energies. The most commonly used radionuclide is ¹⁸F because its relatively long half-life of nearly 110 minutes and its small photonic energy of 0.64 MeV allow greater ease of use and improved resolution, respectively. Glucose is the most commonly used biologic agent because of its relatively ubiquitous use as a form of energy throughout the body. Glucose is labeled with ¹⁸F to create the glucose analog 2-deoxy-2-[¹⁸F] fluoro-D-glucose (¹⁸F-FDG), which is the most widely used radionuclide in oncology because cancer cells have greater metabolic activity compared with most normal cells and exclusively use glucose as a source of energy.

The radionuclides used in PET emit a positively charged electron (positron) as they decay. These positrons are annihilated after encountering an electron and produce a pair of 511-MeV photons that travel in opposite directions, which are then detected by the PET scanner.

Cancer cells are capable of greater intracellular uptake of FDG because of increased glucose transporters on the cell membrane. The glucose is then converted to FDG-6-phosphate, which cannot be further metabolized and remains trapped in the cell. Because of greater accumulation in cancer cells, more positron emission events occur in the tumor compared with surrounding normal tissue. As hundreds of thousands of coincidence events occur, they can be statistically traced back to their origin, allowing spatial localization.

Due to the relatively poor spatial resolution of PET, disease localization often can prove difficult. To circumvent this problem, CT is combined with PET to provide spatially matched morphologic and functional data. PET and CT images can be integrated using three different techniques (Table 1). Although there are advantages and disadvantages to each technique, the integrated PET/CT study using a single machine provides the best coregistration of physiologic and anatomic detail [2]. In the integrated machine, both a diagnostic CT scan and a low-dose transmission scan are obtained. The diagnostic CT scan, often obtained with the administration of contrast material, provides excellent anatomic data.

By creating an attenuation correction map, the transmission CT allows for a reduction in attenuation correction artifacts. These artifacts occur because of the greater attenuation of photons originating deeper within the body or within or adjacent to dense structures, such as bone, compared with those originating from the surface of the body or within or adjacent to less dense structures. The use of the transmission CT for attenuation correction also significantly reduces the PET scanning time by up to 40% compared with a stand-alone PET scanner [3]. Both the diagnostic CT and the transmission CT can be fused with the PET images. Although the anatomic detail is superior in the diagnostic CT scan, the transmission CT scan provides more precise coregistration because both it and the PET are obtained during quiet respiration [4].

The amount of FDG uptake in any region can be visually assessed or quantified by using methods such as the standardized uptake value (SUV). For visual assessment, the up-take in a certain region is subjectively compared with the surrounding background activity. This is often site specific. For instance, uptake in the lung is much less than in the mediastinum. The SUV is a unitless ratio that compares the amount of uptake in a tissue per unit of volume and divides it by a normalizing factor. An SUV of 2.5 is sometimes used to differentiate benign from malignant disease. Although this method is sensitive, it has a very low negative predictive value and should not be the only method used in the decision-making process [5].

PET/CT is the preferred noninvasive method for staging NSCLC. TNM staging [6] plays an important part in determining treatment strategy, which includes surgery, chemotherapy, and radiotherapy. These strategies can be used alone, concurrently, or in an adjuvant or neoadjuvant setting (Table 2).

Determination of the tumor (T) designation is based on the size of the tumor, involvement of contiguous structures, and the presence or absence of satellite nodules (Table 2). T1–T3 tumors are potentially resectable, whereas many T4 tumors are considered inoperable.

PET/CT more accurately determines the T designation compared with either CT or PET alone. In one meta-analysis, PET/CT accurately predicted the T stage in patients with NSCLC in 82% of cases compared with 55% and 68% with PET alone and CT alone, respectively [7]. One of the advantages of PET/CT is in differentiating central tumors from postobstructive atelectasis because the tumor will often have increased FDG uptake compared with an atelectatic lung [8] (Figs. 1A and 1B). PET/CT also improves detection of subtle areas of invasion that may be occult on CT alone [9] (Figs. 2A and 2B). Although PET/CT is an excellent noninvasive method for determining the degree of tumor invasion, pathologic staging through surgical intervention remains the gold standard.

In CT, morphologic characteristics, such as size, are used to predict pathology. A lymph node with a short-axis diameter greater than 1 cm is considered enlarged and a predictor for metastasis. However, this method has proven inaccurate. In one study, 44% of metastatic lymph nodes in patients with NSCLC measured less than 1 cm, and 77% of patients without metastatic lymph nodes had a lymph node measuring greater than 1 cm in the short-axis diameter [10].

PET is more accurate in detecting lymph node metastases compared with CT alone [11]. However, its poor spatial detail can lead to inaccuracies, particularly in areas of normal physiologic uptake (Table 3). In a study by Cerfolio et al. [12], the accuracy of lymph node staging by PET and PET/CT was 56% and 78%, respectively, when compared with mediastinoscopy or surgical staging. By integrating functional and anatomic data, PET/CT is the best noninvasive method for the detection of nodal metastasis [13] (Figs. 3A and 3B), although mediastinoscopy remains the gold standard [14].

NSCLC can metastasize to nearly any organ in the human body, but it most commonly spreads to the brain, liver, adrenal glands, bone, and lung. On initial staging, CT alone can show definitive evidence of metastatic disease in 11–36% of patients [15]. However, the addition of PET reveals occult distant metastases in 5–29% of patients [16]. Nonetheless, the lack of spatial localization can once again lead to errors. By fusing the two datasets together, PET/CT provides an improved method to accurately evaluate for local and distant metastatic disease [1].

Findings such as a small second lung nodule, small adrenal nodule, or subtle bony abnormality are encountered in many patients undergoing cross-sectional imaging. Increased FDG uptake in an adrenal nodule has high specificity and sensitivity in predicting malignancy [17] (Figs. 4A, 4B, and 4C). Similarly, increased uptake in osseous structures can be seen with subtle or no CT abnormality and can suggest metastases [18] (Figs. 5A, 5B, and 5C).

In the newly adopted 7th edition of the TNM classification for lung cancer, NSCLC metastatic disease is divided into local (M1a) and distant (M1b) spread [6] (Table 2). Although the presence of pleural or pericardial nodules on CT can confirm the diagnosis of M1a disease, these findings are often absent. PET can often suggest the diagnosis of malignant pleural disease by showing increased diffuse or focal FDG uptake, but localization to the pleura or pericardium is not always clear (Figs. 6A, 6B, and 6C). Statistically, PET/CT has proven to be superior to either technique in evaluation for metastatic pleural disease [19].

Additional lung nodules are a common finding on PET/CT. Although increased uptake localizing to a contralateral lung nodule can narrow the differential diagnosis, a second biopsy may be necessary to differentiate among a metastatic nodule, a second primary lung cancer, and a chronic inflammatory or infectious nodule.

Although PET/CT is the best tool to date for the noninvasive imaging of NSCLC, numerous pitfalls exist. Imprecise physiologic and anatomic registration, most common adjacent to the diaphragm and heart, can lead to misregistration artifact [4] (Figs. 7A, 7B, and 7C). Metallic objects within the body, such as stents, can cause increased FDG uptake in adjacent tissue because of attenuation-related artifacts, potentially causing false-positive results [7]. Increased FDG uptake is not limited to cancer cells. Many processes with increased metabolic activity, such as infection and inflammation, show increased uptake on PET [14] (Figs. 8A, 8B, and 8C). Similarly, normal physiologic uptake in the brain, brown fat, myocardium, vasculature, liver, and bowel can often mask or be misinterpreted as pathology [20] (Figs. 6A, 6B, 6C, 9A, 9B, and 9C). Although subcentimeter nodules and lymph nodes can show increased FDG uptake, the accuracy of PET is reduced when lesions measure less than 1 cm [21] (Figs. 10A and 10B). It is important to know of these potential pitfalls so errors in staging can be avoided (Table 3).

The accurate staging of NSCLC is an important factor in determining optimal patient treatment. Although various imaging methods can be used for staging, PET/CT has proven to be the best through the integration of both anatomic and physiologic data.

Address correspondence to S. Digumarthy ([email protected]).

This article is available for CME credit. See www.arrs.org for more information.