Computed tomography (CT) is the primary imaging modality for the diagnosis, staging, and follow-up of most thoracic cavity tumors. Fluorine-18 fluorodeoxyglucose positron emission tomography/CT has established itself as a supplementary tool to CT in lung cancer staging and in the assessment for distant metastases of many thoracic tumors. Magnetic resonance imaging is an important adjunctive imaging modality in thoracic oncologic imaging and is used as a problem-solving tool to assess for chest wall invasion, intraspinal extension, and cardiac/vascular invasion. Imaging can facilitate minimally invasive biopsy of most thoracic tumors and is vital in the pretreatment planning of radiation therapy.
Key points
- •
Computed tomography (CT) is the primary imaging modality used in the diagnosis, staging, and follow-up of most thoracic cavity tumors.
- •
Fluorine-18–fluorodeoxyglucose positron emission tomography ( 18 F-FDG PET)/CT has established itself as a supplementary tool to CT in lung cancer staging and in the assessment for distant metastases of many thoracic tumors.
- •
Magnetic resonance imaging is an important adjunctive imaging modality in thoracic oncologic imaging and is used as a problem-solving tool to assess for chest wall invasion, intraspinal extension, and cardiac/vascular invasion.
Introduction
Thoracic tumors, of which lung cancer is the most common, are an important global health issue. Lung cancer is the most common cancer worldwide, with an estimated 1.8 million cases diagnosed in 2012, of which more than half (58%) were in the developing world. Globally, lung cancer is the most common cause of death from cancer, accounting for almost 1 in 5 cancer deaths (1.59 million deaths, 19.4% of total). Although the 1-year relative survival for this malignancy has increased from 37% in 1975 to 1979 to 44% in 2005 to 2008, the overall 5-year survival rate for all stages of lung cancer remains low at 16%. Worldwide, lung cancer rates have peaked for men; but rates for women continue to increase and are closely linked to smoking rates.
Primary mediastinal neoplasms, with the exception of lymphoma, are rare tumors that can arise from any cell precursor. Because of their low incidence and heterogeneity, diagnosis can be difficult. Staging and treatment pathways are often not well defined. This point is particularly true for thymic epithelial tumors, the most common tumor of the anterior mediastinum.
Tumors involving the pleura are largely caused by metastatic disease, whereas only 10% are considered true primary pleural tumors. The most common primary neoplasms include fibrous tumor of the pleura and malignant pleural mesothelioma. These disease processes can manifest as pleural effusion, pleural thickening, or mass, which can be detected on computed tomography (CT) or magnetic resonance imaging (MRI). In particular, the incidence of malignant pleural mesothelioma (MPM) is increasing worldwide and is expected to peak in industrialized countries in 2010 to 2020.
Introduction
Thoracic tumors, of which lung cancer is the most common, are an important global health issue. Lung cancer is the most common cancer worldwide, with an estimated 1.8 million cases diagnosed in 2012, of which more than half (58%) were in the developing world. Globally, lung cancer is the most common cause of death from cancer, accounting for almost 1 in 5 cancer deaths (1.59 million deaths, 19.4% of total). Although the 1-year relative survival for this malignancy has increased from 37% in 1975 to 1979 to 44% in 2005 to 2008, the overall 5-year survival rate for all stages of lung cancer remains low at 16%. Worldwide, lung cancer rates have peaked for men; but rates for women continue to increase and are closely linked to smoking rates.
Primary mediastinal neoplasms, with the exception of lymphoma, are rare tumors that can arise from any cell precursor. Because of their low incidence and heterogeneity, diagnosis can be difficult. Staging and treatment pathways are often not well defined. This point is particularly true for thymic epithelial tumors, the most common tumor of the anterior mediastinum.
Tumors involving the pleura are largely caused by metastatic disease, whereas only 10% are considered true primary pleural tumors. The most common primary neoplasms include fibrous tumor of the pleura and malignant pleural mesothelioma. These disease processes can manifest as pleural effusion, pleural thickening, or mass, which can be detected on computed tomography (CT) or magnetic resonance imaging (MRI). In particular, the incidence of malignant pleural mesothelioma (MPM) is increasing worldwide and is expected to peak in industrialized countries in 2010 to 2020.
Lung cancer
Diagnosis
Many advances have been made in recent decades in the imaging of lung cancer and molecular diagnostics; however, most new patients with lung cancer still have advanced stage disease at the time of presentation. The 5-year survival rate for localized disease is 53%; but only 15% of lung cancers are diagnosed at this early, potentially resectable stage. New targeted treatments of lung cancer have failed to achieve a significant reduction in mortality; therefore, early diagnosis will remain a crucial aim in the battle against this disease.
Histologic subtypes
Non–small cell lung cancer (NSCLC) represents 85% of cases, and small cell lung cancer (SCLC) accounts for approximately 15%. Adenocarcinoma is the most prevalent subtype of lung cancer in the United States, having replaced squamous cell carcinoma as the most common cell type in recent years. The typical appearance of adenocarcinoma of the lung is that of a solitary pulmonary nodule or mass, often peripheral ( Fig. 1 ). Peripheral adenocarcinomas can invade the pleura and grow in a circumferential manner around the lung, sometimes mimicking a malignant mesothelioma.
In 2011, The International Association for the Study of Lung Cancer (IASLC), the American Thoracic Society, and the European Respiratory Society published new guidelines for the classification of lung adenocarcinoma, which placed an increased emphasis on imaging characteristics. The terms bronchioloalveolar cell carcinoma and mixed subtype adenocarcinoma were eliminated. They introduced the pathologic subtypes of adenocarcinoma in situ (AIS) and minimally invasive adenocarcinoma (MIA) to define a subset of patients who would have close to 100% disease-specific survival rates following a complete resection. AIS refers to small adenocarcinomas with purely lepidic growth. If an invasive component of 5 mm or less is present, the term MIA is applied. Both typically appear as ground-glass nodules on CT ( Fig. 2 ) with the presence of a solid component raising suspicion for invasion. In 2013, the Fleischner Society published recommendations for the follow-up of subsolid pulmonary nodules, taking into account the often-indolent growth of subsolid adenocarcinomas and the risk of an invasive component.
Squamous cell carcinoma is now the second most common histologic subtype and accounts for 24% of cases. Squamous cell carcinoma is strongly associated with smoking. Approximately two-thirds of these tumors are centrally located, and they often have associated postobstructive atelectasis or pneumonia. The remaining third are peripheral and can rarely be distinguished from adenocarcinoma based on imaging alone, although squamous cell carcinomas are more likely to be cavitary lesions ( Fig. 3 ). Large cell carcinoma is an uncommon histologic subtype of NSCLC, accounting for less than 3% of all cases, and commonly presents as a large peripheral mass.
The incidence of SCLC has been decreasing for many years, currently representing approximately 15% of all lung cancer cases in the United States. Like squamous cell carcinoma, SCLC usually arises centrally, often a large lobulated mass at time of diagnosis, with frequent mediastinal and hilar invasion ( Fig. 4 ). Distant metastases are common at presentation.
Screening
An effective screening test for the detection of early lung cancer has been a goal of cancer researchers for several decades. Early randomized screening trials using chest radiography raised concerns about the rate of false-positive results and were inconclusive in demonstrating a mortality benefit from screening. On November 5, 2010, the results of the National Lung Screening Trial (NLST) were published, which was the first trial to show that screening could decrease lung cancer mortality. Their results demonstrated a significant reduction in the mortality rate from lung cancer (20%) in patients who were screened annually with low-dose CT (LDCT). They also found a 6.7% reduction in mortality from any cause.
In July 2013, the US Preventative Services Task Force published new guidelines for lung cancer screening, concluding that there was now strong evidence that LDCT screening can reduce lung cancer and all-cause mortality. Their new guidelines recommend annual screening with LDCT in adults aged 55 to 80 years who have a 30 pack-year smoking history and currently smoke or have quit within the previous 15 years.
LDCT chest for lung cancer screening is performed without intravenous contrast. The low-dose technique results in increased image noise ; but newer reconstruction algorithms are now available for CT, which can improve the image quality of low-dose studies, including adaptive statistical image reconstruction, iterative reconstruction in image space, and the newer model-based iterative reconstruction (MBIR) techniques. The use of MBIR in particular could potentially allow diagnostic quality scans to be performed at even lower doses than are currently being used for screening, but this will require validation with larger studies in the future.
Evaluation of pulmonary nodules
Despite the high spatial and contrast resolution offered by CT, pulmonary nodules are still often missed, for reasons including small size, subsolid composition, location adjacent to vessels, and adjacent parenchymal disease. Computer-aided detection (CAD) is increasingly recognized as a method for increasing the detection of small pulmonary nodules on CT and has been shown to have good sensitivity in the detection of small lung tumors, including those missed on the initial read by a radiologist. However, most lung CAD systems are optimized for the detection of solid, spherical nodules; the sensitivity of these systems for the detection of subsolid nodules remains poor.
Although lung CAD can improve the detection of pulmonary nodules on CT, there remains a need to develop more robust strategies to help differentiate benign and malignant pulmonary nodules, as most small pulmonary nodules are benign. The current follow-up guidelines for solid nodules, as recommended by the Fleischner Society, are based solely on size criteria ( Table 1 ). The high false-positive rate of screening CT chests remains a concern, as it often leads to unnecessary further imaging studies, biopsies, or surgery. In the NLST trial, only 3.6% of the positive CT studies for pulmonary nodules were subsequently proven to represent lung cancer.
| Nodule Size (mm) | Low-Risk Patients | High-Risk Patients |
|---|---|---|
| ≤4 | No follow-up needed | CT at 12 mo; if unchanged, no further follow-up |
| >4–6 | CT at 12 mo; if unchanged, no further follow-up | Initial follow-up CT at 6–12 mo, then at 18–24 mo if no change |
| >6–8 | Initial follow-up CT at 6–12 mo, then at 18–24 mo if no change | Initial follow-up CT at 3–6 mo, then at 9–12 and 24 mo if no change |
| >8 | Follow-up CT at around 3, 9, and 24 mo; dynamic contrast-enhanced CT, PET, +/− biopsy | Same as for low-risk patients |
Fluorine-18–fluorodeoxyglucose positron emission tomography ( 18 F-FDG PET)/CT is an important tool in the assessment of pulmonary nodules but should not be relied on when the nodule size is less than 8 mm. Overall, the lack of FDG uptake in a pulmonary nodule on 18 F-FDG PET/CT predicts a less than 5% chance of malignancy. The degree of FDG uptake within a malignant nodule may also have implications for prognosis.
MRI in lung cancer diagnosis
MRI, because of its lower inherent spatial resolution, is not usually used in the evaluation of pulmonary nodules. Studies have suggested that, although the sensitivity of MRI in nodule detection is lower than that of CT, it is more sensitive in detecting malignant than benign nodules ; MRI diffusion-weighted imaging may be of use in differentiating benign from malignant lesions.
Tissue diagnosis
Percutaneous CT-guided biopsy is the most commonly used method for obtaining a sample for tissue diagnosis of a lung mass. The diagnosis yield for a percutaneous needle biopsy is 36% to 84%, depending on factors such as the size of the lesion, technique used, and lesion location. The sensitivity of a CT-guided biopsy is 94%, with a specificity of 99%. Flexible bronchoscopy with transbronchial needle aspiration is an alternative method for obtaining a biopsy specimen but is best suited for lesions in a central location or those with endobronchial extension. Navigational bronchoscopy uses standard CT images to create a virtual 3-dimensional (3D) bronchoscopy, which can assist in locating lesions in the lung periphery during flexible bronchoscopy.
Staging
Accurate staging in lung cancer is vital for determining appropriate clinical management. Imaging studies play a key role. The seventh edition of the TNM staging handbook, published in 2009, made substantial changes to the staging of NSCLC. The IASLC proposed a new lymph node map in 2009 ( Fig. 5 ), which grouped lymph node stations into zones. The now ubiquitous use of multidetector CT for the staging of lung cancer has been supplemented in the past 2 decades by 18 F-FDG PET/CT, and it is well recognized that the use of newer imaging modalities can supplement CT findings ( Table 2 ) and has resulted in clinically relevant stage migration for patients.
| Modality | Uses |
|---|---|
| CT | Preferred modality for assessing tumor size, location of tumor, proximity to other structures, and presence of pulmonary metastasis |
| 18 F-FDG PET/CT | Mediastinal nodal staging |
| Evaluation for distant metastases | |
| MRI | Assessment of superior sulcus tumors for brachial plexus and subclavian vessel invasion |
| Assessment of vertebral body invasion | |
| Evaluation for brain metastases |
The current guidelines of the National Comprehensive Cancer Network for staging of both NSCLC and SCLC recommend a CT scan (preferably contrast enhanced) of the chest and upper abdomen for initial staging. An integrated 18 F-FDG PET/CT scan is recommended for all patients with NSCLC and for suspected limited stage patients with SCLC to assess for nodal and distant metastases. Integrated 18 F-FDG PET/CT has been found to be more accurate for staging NSCLC than 18 F-FDG PET alone, CT alone, or in comparison with separate PET and CT studies. The positive predictive value of 18 F-FDG PET/CT for nodal metastases is better than that of CT (80%–90% compared with approximately 50%).
When the 18 F-FDG PET/CT findings suggest nodal metastases in NSCLC, pathologic confirmation is recommended by mediastinoscopy, mediastinotomy, endobronchial ultrasound (US), or CT-guided biopsy. In patients with NSCLC who are clinically T1N0, the absence of nodal uptake on 18 F-FDG PET obviates histologic correlation with invasive mediastinal node sampling.
CT has some limitations in the staging of lung cancer, particularly in assessing for chest wall or mediastinal invasion. The role of MRI in staging of lung cancer is small, but it can be used to assess for vascular or cardiac invasion or invasion of vertebral bodies in suspected T4 tumors. MRI of the brachial plexus is occasionally useful when staging Pancoast tumors ( Fig. 6 ). Potential future uses for MRI in lung cancer staging include the use of diffusion-weighted imaging to assess for nodal metastases, which several studies have suggested is superior to 18 F-FDG PET/CT in nodal staging. MRI is the preferred imaging modality to assess for brain metastases.
Treatment
Despite many recent advances in treatment of primary lung cancer, surgery remains the current standard of treatment of early stage disease. Recently, nonsurgical treatments, including stereotactic body radiation (SBRT) for stage I NSCLC have become more commonly used, particularly in patients who are poor candidates for surgery. In SBRT, several fractions of very-high-dose radiation therapy are administered to small lung tumors, which must be remote from the central mediastinal vessels, esophagus, and main bronchi. The use of SBRT requires careful pretreatment planning with respiratory-correlated CT scans and image guidance of radiation delivery. Accelerated or hyperfractionated radiotherapy regimes have also been proven to confer a survival advantage over conventional radiotherapy regimes in patients with locally advanced lung cancers. 18 F-FDG PET/CT is an important modality in the pretreatment evaluation of radiotherapy; its use can result in smaller subsequent radiation fields, potentially allowing for radiation dose escalation with fewer side effects. 18 F-FDG PET/CT integration has also been proven to reduce the variability among radiation oncologists in delineation of the primary tumor and can facilitate automatic tumor delineation.
Radiofrequency (RF) ablation is an alternative treatment option for early stage tumors in patients who are not suitable candidates for surgery and can also be used in the treatment of lung metastases as well as in the palliative treatment of chest wall masses. Pretreatment imaging in RF ablation helps to guide access trajectory and can also influence probe choice and the use of adjunctive imaging modalities, such as US and CT fluoroscopy.
Follow-up
The evaluation of tumor response to nonsurgical therapy has become an important topic in recent years, particularly since the advent of targeted therapies to specific molecular targets in lung cancer. CT has traditionally been the most widely used diagnostic tool in assessing tumor response, but the utility of molecular imaging is becoming increasingly recognized because of its potential to characterize tumor tissue by its biochemical and biological characteristics. Alterations in cellular metabolism usually precede change in tumor size, and the metabolic response on PET has been shown to correlate better with outcome than size response on CT. In addition to monitoring for the response to standard chemotherapy, PET can also be used to assess the response to biological therapies, such as the epidermal growth factor receptor (EGFR) kinase inhibitor erlotinib. Recent studies have shown that 18 F-FDG PET/CT can separate metabolic responders from nonresponders as early as 2 weeks following initiation of targeted therapy.
Evolving techniques with potential for clinical use include the development of new molecular tracers targeting different aspects of tumor biology, such as 18 F-fluorothymidine, and the assessment of EGFR and EGFR tyrosine kinase overexpression in tumors by PET imaging. In vivo a priori determination of the efficacy of EGFR-targeted drug therapy is a promising imaging tool for the future and may ultimately help in the development of individualized treatment plans for patients. Posttreatment tumor response has also been investigated using various MRI techniques, such as diffusion-weighted and perfusion imaging and magnetization transfer.
Newer nonsurgical treatment modalities, such as RF ablation and SBRT, have made it essential for the radiologist to be familiar with the expected, treatment-attributable changes in the lung parenchyma on follow-up imaging. Immediately following RF ablation, a region of focal ground-glass change surrounds the lesion on CT. This ground-glass change usually resolves within a month. Occasionally, these lesions cavitate but with a gradual decrease in size of the cavity on subsequent follow-up. On postablation 18 F-FDG PET/CT, a rim of FDG-avid inflamed tissue is often seen surrounding the treated lesion. This uptake may persist for several months. However, increasing FDG uptake within the ablated lesion over time is suspicious for local recurrence of tumor rather than posttreatment inflammation.
Radiation-induced lung changes after SBRT differ from those seen following conventional radiotherapy. Changes in CT lung density, particularly masslike consolidation, are common post-SBRT and can cause overestimation of tumor recurrence. Distinguishing between radiation-induced changes and local recurrence is of paramount importance. The current practices mainly rely on CT for follow-up, but the adjunct use of 18 F-FDG PET/CT is increasingly recognized where local recurrence is suspected, although its use remains controversial.
Careful imaging follow-up is required in all patients with lung cancer to detect local recurrences at an earlier stage in the initial years following treatment and to evaluate for the development of a second primary lung tumor in later years. The risk of developing a second primary lung cancer is substantial: approximately 1% to 2% per year following resection of an NSCLC and 6% in patients who have survived SCLC. The risk increases to greater than 10% per patient per year 10 years following the initial treatment of SCLC. Death from a second primary tumor is, thus, common in lung cancer survivors. The current recommendations for follow-up after the treatment of NSCLC recommend a contrast-enhanced CT scan every 4 to 6 months for the first 2 years to optimally evaluate for mediastinal recurrence, followed by a yearly non–contrast-enhanced CT scan to assess for new lesions in the lung parenchyma.
Other Primary Lung Tumors
Bronchopulmonary carcinoid tumors are relatively rare malignant tumors, most indolent in behavior, but with variability depending on tumor grade. All have the potential for invasion and metastases. Approximately 25% present as a central mass, often involving a lobar or segmental bronchus ( Fig. 7 ). Peripheral carcinoids are generally well-circumscribed, homogeneous, slow-growing masses. A contrast-enhanced CT scan is essential for the evaluation of a central carcinoid tumor, to assess the relationship with hilar structures, and to identify lymphadenopathy suggesting nodal metastases. The role of 18 F-FDG PET/CT is not well established in the imaging of carcinoid tumors and has a wide range of quoted sensitivity (14%–100%). One-third of carcinoid tumors are somatostatin-receptor negative; the remainder exhibits only weak uptake on octreotide scans, so routine use of octreotide scanning is not recommended.
