Imaging in Lung Cancer Staging: Overview, TNM Staging, T Classification of Primary Tumor
Imaging modalities are valuable in defining the T classification of the primary tumor by providing information about its size and local extent/invasion. Chest CT scanning has been the primary and most commonly used modality to achieve this task. This role has been enhanced since the introduction of multidetector CT (MDCT) scanning. This is primarily because of the ability of MDCT to generate thin sections in a shorter scan time, resulting in improvement of the spatial resolution and multiplanar reformation (MPR) capability. 
MRI and PET scanning provide additional crucial information for accurate classification in some cases.  Compared with CT scan, MRI has superior tissue contrast but it is more susceptible to cardiac and respiratory motion artifacts. It is also affected by low proton density, very short T2* values, and inhomogeneity of the magnetic field in the lungs. However, recent advances in MRI techniques and the use of gadolinium contrast media have enhanced the diagnostic capability of MRI in detecting and staging lung cancer.  MRI can be used to further delineate the findings of nonenhanced CT in patients with adverse reactions to iodinated contrast media or significant renal impairment. 
Chest radiography and ultrasound have limited utility in the staging of lung tumors.
The new staging system does not provide specific guidelines for the imaging-based measuring technique of the primary neoplasm. The maximal dimension of the lesion on multiplanar reconstructed CT scans (in lung window settings) is generally used.  In tumors with surrounding hemorrhage or changes of postobstructive pneumonitis and/or atelectasis, estimating the size of the primary neoplastic lesion by CT scanning can be challenging (especially on noncontrast CT). In some of these cases, PET scanning and MRI can delineated the tumor from the surrounding changes.
Chest radiography can demonstrate the size of the primary tumor, especially in peripheral lesions. However, its utility is limited in cases of central lesions or those with surrounding secondary processes. Additionally, chest radiography is less accurate than CT scanning in estimating the size of the peripheral tumors, owing to magnification effect and issues related to projection.
In clinical practice, the endobronchial extent of the primary neoplasm is primarily and accurately defined by bronchoscopy. Imaging, especially CT scanning and MRI, can suggest the extent of the tumor by delineating its endobronchial component or by demonstrating associated secondary changes such as postobstructive pneumonitis or atelectasis.
Visceral pleural invasion (VPI) is an important finding that advances the classification from T1 to T2. In the seventh edition of the TNM staging, the IASLC implemented a standardized definition of the visceral pleural invasion as extension beyond the elastic layer up to the visceral pleural surface (based on pathologic examination with elastic stains).  Invasion of the pleura into an adjacent ipsilateral lobe should be classified as T2. Imaging is limited in reliably confirming invasion in the presence of contact between the mass and the visceral pleura.
Recently, measurement of the ratio of the interface between the tumor and adjacent structure (arch distance) to the maximum tumor diameter has been suggested as a high performance criterion of pleural invasion.  A ratio greater than 0.9 achieved sensitivity and specificity of 89.7% and 96%, respectively.
The invasion of the chest wall can be reliably detected by CT only in cases with gross soft tissue or osseous involvement. Otherwise, CT is limited in assessing such invasion in less overt cases with contact between the mass and the chest wall. Detecting chest wall invasion by lung tumor is important for preoperative evaluation and planning. Although such invasion does not exclude surgery, it may alter the surgical approach. The following CT findings can suggest but do not confirm the invasion of the parietal pleura and chest wall:
Length of contact between the lesion and the chest wall of more than 3 cm
Relatively large area of contact between the mass and the chest wall, with the ratio of the lesion's diameter to the length on contact of more than 0.5
of the extrapleural fat plane
Obtuse angle and/or pleural thickening at the margin of contact between the lesion and the pleura
MRI is more accurate than CT scanning in depicting chest wall invasion, especially in the superior sulcus region. However, the accuracy of both conventional CT and MRI in assessing chest wall invasion is relatively low. Emerging ultrasound, CT, and MRI dynamic techniques that assess the tumor movement in relation to the chest wall could be helpful in the confirmation of chest wall invasion but are not widely implemented. In a 2008 study,  ultrasound was more sensitive than CT in evaluating chest wall invasion, with sensitivity of 89% versus 42%. In another 2008 study about chest wall invasion detection by respiratory dynamic MRI,  the sensitivity and specificity were 100% and 82.9%, respectively. The false-positive cases in these 2 studies were mostly secondary to benign adhesions between the visceral and parietal pleura.
Plain radiography has a limited role in detecting chest wall invasion, except in the presence of large soft tissue mass or osseous destruction.
Mediastinal invasion by the primary tumor can be reliably depicted using CT scanning or MRI. This is one of the important criterion in staging lung cancer, since it can determine resectability of the tumor.
In some cases, mediastinal invasion can be grossly visible (eg, significant replacement of the fat by the soft tissue of the tumor or encasement of a vascular structure or other structures by the neoplasm). Otherwise, the contiguity of the tumor with the mediastinum does not necessarily imply invasion. It was suggested that although CT cannot confirm mediastinal invasion in such cases with certainty, it can separate masses that were likely to be technically resectable (no invasion or focal limited invasion).  This separation of the resectable lesions is based on the presence one or more following CT findings:
Contact between mass and mediastinum of less than 3 cm
Circumferential contact between the mass and aorta of less than 90°
Presence of a fat plane between the mass and the mediastinal structures
Using the reverse criteria is less reliable in predicting irresectability. For example, a fat plane between the mediastinum and mass can be obliterated not only by invasion, but also by technical factors (eg, motion artifacts and volume averaging). The presence of more than 3 cm of contact or more than 90° of encasement has low sensitivity in confirming invasion.
MRI can be used in cases with questionable CT findings, given its superior tissue contrast and multiplanar capabilities. In a study from 2013, the sensitivity in nonspecific CT, CT with contrast-enhancement, and MR angiogram for detecting mediastinal and hilar invasion where 78-90%, 73-87%, and 75-88%, respectively.  MRI can delineate infiltration or disruption of the extrapleural fat planes, which suggests chest wall invasion. This can be further enhanced by the administration of intravenous contrast and other nonemergent techniques such as dynamic cine MRI.
The combined PET/CT is more accurate than PET alone in detecting chest wall and mediastinal invasion. However, there is limited added value of PET/CT over CT in T assessment. 
The presence of additional separate tumor nodule(s) renders the tumor's classification as T3 if the nodule is in the same lobe of the primary tumor, or as T4 if the nodule is in a different ipsilateral lobe. A separate tumor nodule in the contralateral lung is considered M1a. CT is the primary modality to evaluate for additional tumor nodules. The differential diagnosis of a separate nodule on CT should include benign entities and synchronous primary tumor in addition to metastasis from the primary lesion. Synchronous primary tumors, which should have different histologic cell types or subtypes, are not considered as T4 but are classified according to the highest the T designation of the lesions with a number of nodules in parentheses. PET CT can further suggest a malignant nature of an additional nodule by demonstrating increased fluorodeoxyglucose (FDG) uptake.