CONTRIBUTION OF MRI IN LUNG CANCER STAGING*

Major advances in the WB­MRI in the initial evaluation and follow­up of patients with lung cancer have been per­ formed in recent years. Multicentric studies using different magnet systems are necessary to confirm these promising results.

Lung cancer is the leading cause of cancer-related death worldwide, with a dismal 5-year survival rate of 15% (1). It accounts for 12.2% of all new cases of cancer in Europe in 2008 (1) and 14% of all new cases of cancer in the USA in 2011 (2). Accurate staging is mandatory to select the most appropriate therapy and to determine prognosis.
The two advanced imaging methods used for this staging were CT-scan and 18 F-FDG PET/CT. However, both technics had some limitations. Limitations of 18 F-FDG PET/CT are particularly limited spatial resolution (3) and low specificity in distinguishing malignant nodule or lymphadenopathy from inflammatory changes, resulting in a considerable number of false-positive findings (4). The 18 F-FDG PET/CT is not recommended for brain staging. Moreover, PET/CT is associated with a considerable radiation burden to patients and medical personnel. Limitations of CT-scan are the use of morphological (aspect and size of the nodule, size of the small diameter of the lymphadenopathy) data without functional or biological information.
Magnetic resonance imaging (MRI) is currently the only technique that enables non-invasive wholebody assessment without ionizing radiation. Another strength of MRI is its capability to create high soft tissue contrast without external contrast agents and with high spatial resolution. Currently, MRI is recommended in the assessment of lung cancer extension to the lung apices (superior sulcus or Pancoast-Tobias tumor), to the spinal cord, and to the cardiac cavity. For metastases issues, MRI is also recommended for its high sensitivity and specificity for brain, bone, liver and adrenal metastases diagnosis. Recent advances in Radiologist is more confident with MRI for extension evaluation of the tumors within the superior sulcus.
A B D C compared (10)(11)(12). MRI in all these studies was superior to CT for the assessment of brachial plexus invasion related to the multiplanar MR imaging and contrast resolution. No data are available comparing multidetector CT and MR (7).
MR imaging of superior sulcus tumors is performed by using a protocol described by Bruzzi et al. (7) using a modification of a previous protocol described by Demondion et al. (13,14). This protocol includes axial, sagittal, and coronal T1weighted sequences and sagittal T2weighted sequence. To optimize sensitivity for the small structures in the thoracic inlet, such as the brachial plexus nerve roots and trunks, imaging is performed by using a neurovascular neck coil. T1-weighted sequences are acquired by using thin section (3.0 mm) with a minimal gap (< 0.3 mm) and both cardiac gating and respiratory triggering are used to minimize motion and pulsation artifact. Sagittal T1-weighted sequences provide the most detailed anatomic information and should be performed first in case imaging has to be interrupted or aborted, because the sagittal images alone may provide sufficient diagnostic information.
the characterization of suspected lesions of liver and adrenal.

Superior sulcus tumor
MRI advantages in the evaluation of superior sulcus tumor and determining their resectability include multiplanar capabilities, superior contrast resolution (compared with the other modalities), and lack of ionizing radiation (Fig. 1). MRI is superior to CT in the visualization of tumor extension to the chest wall, extending into the foramina and spinal canal, and the involvement of the brachial plexus (8)(9)(10)(11). Although tumor invasion of these structures can be inferred scan data in many cases, MRI allows direct representation of participation and thus improve reader confidence (10). Disadvantages of MRI include its limited availability compared to that of CT, as well as longer time image acquisition and increased sensitivity to motion artifacts and patient claustrophobia. There have been a limited number of prospective studies, conducted in the late 1980s and early 1990s, with small number of patients in which the relative merits of CT and MR imaging of the superior sulcus were MRI are made around the tumor functional exploration including a whole body exploration. This functional exploration focuses on the specific cellular and vascular architecture of tumors using MRI spectroscopy, perfusion MRI and diffusion-weighted images (DWI). The image contrast of DWI is based on the diffusion properties of water molecules and reflects tissue parameters like cellular density especially in tumor and tissue architecture (5). In the last few years, DWI has been investigated successfully in many fields of oncology (6).
In this review we present the contribution of MRI in lung cancer staging including the validated indications and the current development especially with DWI.

validated indication of MRI in lung cancer staging
In the initial staging, MRI is the gold standard in the detection of brain metastases and the chest wall invasion especially of the superior sulcus tumor (7). It is also recommended in cases of suspected vertebral or epidural localization and in example, when a patient is being evaluated for curative liver resection (19). The diagnostic performance of DWI is equal to that of Gd-MRI. DWI alone can be used in patients where gadolinium contrast administration is not allowed. Combination of Gd-MRI and DWI significantly increases diagnostic accuracy (20).

Bone metastases
MRI is both sensitive and specific for diagnosing skeletal metastases (Fig. 3), and previous limitations have been overcome with the introduction of whole-body MRI (21,22).

Adrenal metastases
The discovery of an adrenal gland mass more than 5 cm in the context of lung cancer most often corresponds to a metastatic lesion, except choice (17). It has particular advantages in showing lesions in the posterior fossa and adjacent to the skull. Given its overall higher sensitivity, MRI is therefore currently preferred over CT when screening patients with lung cancer for brain metastases.

Liver metastases
MR imaging with gadolinium chelates offers an accurate non-radiation based imaging test for detection of liver metastases (18). Liver specific MR contrast agents (hepatobiliary and reticuloendothelial agents) offer greater lesion-to-liver contrast than the conventional extracellular agents (gadolinium chelates). Liver specific MR contrast agents may be used in selected clinical situations when the goal is to achieve the highest detection rate for liver focal lesions, for Indications of contrast medium are in patients in whom vascular invasion or intraforaminal extension is suspected to be present (Fig. 2); in patients who have undergone neoadjuvant therapy before a planned resection, in whom posttreatment fibrosis may result in blurring of the intermuscular fat planes and difficulty in visualizing the primary tumor; and in patients in whom a recurrence is suspected after definitive treatment (7).

Brain metastases
MRI of the brain is more sensitive and may be more specific for metastases than CT (15,16). Cerebral metastases occur commonly in lung cancer, particularly from poorly differentiated tumors and adenocarcinoma. MRI with contrast enhancement is the image technique of A B C D cols, as it can be performed relatively quickly (as short as two breath-hold acquisitions or during free breathing or with respiratory triggering) and does not require contrast agent injection, which makes it attractive in patients with decreased renal function, who cannot receive gadolinium-based contrast agents. This recent development of DW leads to other promising opportunities than the detection and characterization of pulmonary lesions, such as the initial staging of lung cancer with the TNM staging and to monitor treatment.

Tissues characterization
Tissue characterization in lung nodule or mass, likes other organs, stays a challenge even with the development of 18 F-FDG PET/CT and the kinetic contrast enhancement using CT or MR. Some DWI MR studies focus on tumor detection and characterization of lung nodules or aging when the protons are in phase. In contrast, nonadenomas do not show signal loss on out-of-phase imaging (23). Recent studies have shown that 60 to 89 percent of lesions measuring between 10 and 30 HU on unenhanced CT can be characterized using chemical shift MRI (24,25).

Current development of MR in lung cancer
Powered by tremendous advances in image quality over the past few years, diffusion-weighted imaging with or without background signal suppression has drawn strong interest from the radiologic community and major MR vendors. DW imaging is increasingly used in the thorax, particularly in lung nodules and masses, with promising results for lung nodule lesion detection and characterization. DW imaging can be easily implemented in clinical proto-myelolipoma and adrenal cyst, the characteristics of their content, with fat for first and liquid for the second, are easily identified. The problem is especially for small lesions less than 3 cm and having a density enhancement after injection. In most cases, insofar as it is an initial assessment, no previous review is available. The problem of finding these adrenal lesions can be studied by the structural approach in differentiating benign and malignant lesions on the basis of the presence or absence of intracytoplasmic lipids. In benign lesions, lipids are observed, whereas in malignant lesions, the cells containing them are destroyed (Fig. 4). Chemical shift MRI uses a technique based on hydrogen and fat protons, which resonate at different frequencies. By using different time parameters during the same MRI examination, it is possible to identify lipid-rich adenomas. These adenomas show signal loss on outof-phase imaging, as opposed to im-

Mediastinal and hilar nodal staging
In patients with NSCLC, involvement of the mediastinal lymph nodes is an important prognostic factor because accurate disease staging is needed to limit surgery or multimodality treatment to only of those who might benefit from such treatment (Fig. 7). A recent meta-analysis (36) comparing 18 F-FDG PET/CT to DWI showed that DWI has a high specificity for N staging of NSCLC compared with 18 F-FDG PET/CT and has the potential to be a reliable alternative noninvasive imaging method for the preoperative staging of mediastinal and hilar lymph node in patients with NSCLC (37-41). However, they believe it is too early to call for broad application of this method in clinical practice. They speculate that additional improvement of the technology will increase its role in the future. Additional, larger, prospective, directly comparative studies involving 18 F-FDG PET/CT would be required to determine the true value of DWI for the diagnosis of lymph node metastasis in patients with NSCLC.
DWI has its place in some special situations to reduce the failure of transthoracic biopsy for large partially necrotic masses (direct biopsy area to which the cell density is highest, the lowest ADC) and differentiates atelectasis from tumor to show the target biopsy (Fig. 6). In the latter situation the DWI is more accurate than other sequences including T2weighted normal (35). MRI always keeps a place in the characterization of silicotic nodules of patients exposed to silica with no signal on T2-masses (26)(27)(28)(29)(30)(31)(32)(33) (Table I). MR is as accurate as 18 F-FDG PET/CT for nodule or masses characterization (Fig. 5). MR is more specific comparing to 18 F-FDG PET/CT in characterization of lung masses or mediastinal lymphadenopathies. In a recent meta-analysis on nodule or mass characterization Wu et al, confirm this assessment on showing that DWI is useful for differentiation between malignant and benign pulmonary nodules with pooled sensitivity of 0.84 and specificity of 0.84. Large-scale randomized controlled trials are still necessary to assess and confirm its clinical value. A threshold value for malignant/benign lesion classification could not be made based on this study because it is influenced by different b values, bias of patient selection, lesions' pathological characteristics and ADC measurement. Selection of the threshold value should be determined according to the purpose of examination. A relatively higher threshold value may be recommended to minimize missing malignancy in lung cancer screening. If DWI is appended to other diagnostic method (e.g., computed tomography), a relatively lower threshold value may be recommended to reduce false-positive results.    In a more recent study of Chen et al. (42), 62 lesions were considered as metastases based on initial findings, 37 distant metastatic lesions (brain, three; liver, six; adrenal gland, two; and bone, 26) and six lung metastatic lesions were validated by biopsy or radiologic follow-up. A total of 35 distant metastases were detected based on DWI. Three lesions of lung metastases, sized less than 10 mm, were not detected at DWI; there was one false-positive bone lesion with DWI. Meanwhile, 37 distant metastases were detected with 18 F-FDG PET/CT; five lung metastatic lesions were detected by 18 F-FDG PET/CT. Only one lung metastatic lesion was missed and no false-positive result at 18 F-FDG PET/CT. DWI was found to be sensitive in osseous metastasis. The sensitivity, specificity, positive predictive value, negative predictive value and accuracy for detection of metastasis for DWI than did 18 F-FDG PET/CT (39), but also DWI gave fewer false-negative results for N staging of NSCLC than did 18 F-FDG PET/CT. 18 F-FDG PET/CT is likely to show false positive results when lymph nodes contain inflammation and is likely to show falsenegative results when the lymph nodes contain a small amount of cancer cells. The DWI with an ADC Nomori et al. (39) reported that the accuracy of N staging in 88 patients was 0.89 with DWI, significantly greater than the value of 0.78 obtained with 18 F-FDG PET/CT because of less overstaging in the former. The superiority of DWI can be explained by the observation that not only did DWI give fewer false-positive results for N staging of NSCLC A B D C net systems are necessary to confirm these promising results. One thing is certain, for metastatic and lymph node staging, that whole-body MRI with the information obtained by WB-DWI and WB-MRI is greater than the scanner including staging and

Conclusion
Major advances in the WB-MRI in the initial evaluation and follow-up of patients with lung cancer have been performed in recent years. Multicentric studies using different mag-were 0.9; 0.95; 0.97; 0.83 and 0.92 respectively. The sensitivity, specificity, positive predictive value, negative predictive value and accuracy for detection of metastasis for integrated 18 F-FDG PET/CT were 0.98, 1; 1; 0.95 and 0.98 respectively.