FormalPara Key Points

Since the publication of the Radiologic Diagnostic Oncology Group Report (RDOG) in 1991, the clinical application of pulmonary magnetic resonance (MR) imaging to patients with lung cancer has been limited. MDCT has been much more widely available for staging of lung cancer in clinical situations. Traditionally, FDG-PET or PET/CT is the only modality that reveals biological glucose metabolism of lung cancer, and ventilation and/or perfusion scintigraphy is the only modality that demonstrates pulmonary function. However, recent advances of MR systems and utilization of contrast media make it possible to overcome the limitation of chest MR imaging. Therefore, in the last years, several investigators have demonstrated the significant comprehensive potential of MR imaging to substitute for MDCT and nuclear medicine examinations in lung cancer staging. Currently, MR imaging in lung cancer patients can be applied for (1) detection of pulmonary nodules, (2) characterization of solitary pulmonary nodules, and (3) assessment of TNM classification in routine clinical practice. We believe that further basic studies, as well as clinical applications of newer MR techniques, will play an important role in the future management of patients with lung cancer including MRI.

1 Introduction

Lung cancer is the most common cause of cancer-related death in the Western world, Japan, and South Korea. Non-small cell lung carcinoma (NSCLC) accounts for approximately 80% of all lung cancers, with small cell lung carcinoma (SCLC) accounting for the remainder. Despite major efforts aimed at improving survival during recent years, survival remains dismal at 14% for all stages. Imaging techniques currently are essential for the diagnosis, staging, and follow-up of patients with lung cancer. The diagnosis of lung cancer has relied on findings on chest radiographs (CXRs) and detection of cells in sputum or biopsy specimens. Perhaps even more important, however, are specific findings on chest computed tomography (CT) and metabolic information on positron emission tomography with 2-[fluorine-18]-fluoro-2-deoxy-d-glucose (FDG-PET) or FDG-PET co-registered or integrated with CT (co-registered or integrated FDG-PET/CT). Moreover, the staging and follow-up of lung cancer have relied more on CT, FDG-PET, and/or FDG-PET/CT than on chest radiography. The goal of diagnosis and management of pulmonary nodules is to bring promptly to surgery all patients with operable malignant nodules while avoiding unnecessary thoracotomy for patients with benign lesions. It is therefore of utmost importance to differentiate malignant from benign nodules in the least invasive manner and to make as specific and accurate a diagnosis as possible. In addition, the preliminary goal of pre-therapeutic assessment of lung cancers is to avoid unnecessary surgery for patients with locally unresectable tumors and/or nodal or metastatic disease because the strongest prognostic indicator for survival in lung cancer is whether or not the tumor is resectable.

Currently, CT is considered to be the most widely applicable modality for evaluation of lung cancer, and a major breakthrough in CT technology has been the introduction of multidetector-row CT (MDCT), in which detector rows are exposed simultaneously. The performance of MDCT compared with single-detector CT is enhanced by a factor approximately equal to the number of rows. In addition, FDG-PET or PET/CT qualifies as another important innovation in lung cancer imaging. Standard imaging techniques are based on differences in the structure of tissues, whereas FDG-PET or PET/CT can show the enhanced glucose metabolism of lung cancer cells. For these reasons, MR imaging has been utilized for only a few cases, such as superior sulcus (Pancoast’s) tumor, mediastinal invasion and chest wall invasion since 1991, when the Radiologic Diagnostic Oncology Group (RDOG) reported no significant difference in the diagnostic capability of CT and conventional T1-weighted imaging for the staging of lung cancer except for mediastinal invasion (Webb et al. 1991). However, recent advances in MR imaging techniques and utilization of contrast media have resulted in further improvement of the image quality and diagnostic capability of MR imaging for lung cancer patients.

In this chapter we describe the utility and capability of MR imaging for (1) detection of pulmonary nodules, (2) characterization and management of pulmonary nodules, and (3) assessment of tumor-node-metastasis (TNM) stages in lung cancer patients.

2 Detection of Pulmonary Nodules

A pulmonary nodule is radiologically defined as an intraparenchymal lung lesion that is less than 30 mm in diameter and is not associated with atelectasis or adenopathy (Tuddenham 1984). While one in 500 CXRs shows a lung nodule, 90% of these nodules are incidental radiological findings, detected accidentally on CXRs obtained for unrelated diagnostic workups. More than 150,000 patients per year in the United States present their physicians with the diagnostic dilemma of a pulmonary nodule. This number has increased even further due to incidental findings of lung nodules on chest CT (Tuddenham 1984). The devastating effect of lung cancer is directly associated with its delayed presentation. Patients with the best prognosis are those with stage IA disease, although approximately one half of all lung cancers unfortunately show extrathoracic spread at the time of diagnosis. Timely and accurate detection and diagnosis of the etiology of pulmonary nodules are therefore essential for making it possible for patients with malignancy to be cured.

Spiral CT or MDCT can be considered the current gold standard for the detection of lung nodules (Davis 1991; Costello 1994; Henschke et al. 2001; Schaefer-Prokop and Prokop 2002; Swensen et al. 2005; Bach et al. 2007). However, repeated follow-up CT examinations for detection of pulmonary metastases may be undesirable, especially for young patients, because of radiation exposure. Although radiation exposure is usually no major issue for cancer patients and low-dose CT techniques have been proved feasible to reduce the radiation dose, MR imaging does not require any ionizing radiation at all. It would therefore be helpful if MR imaging could be used for the detection of pulmonary nodules without administration of contrast media.

Several investigators have addressed this issue by using various sequences with 1.5 and 3.0 T scanners since 1997. However, patient-related motion artifacts, susceptibility artifacts from the lungs, and inferior spatial and temporal resolution as compared with those of CT reduce the quality of MR images of the lungs (Kersjes et al. 1997; Vogt et al. 2004; Schroeder et al. 2005; Luboldt et al. 2006; Bruegel et al. 2007; Regier et al. 2007; Yi et al. 2007). All these studies assessed the detection rate (sensitivity) for pulmonary nodules, mainly pulmonary metastases, which was verified by single helical CT or MDCT. The sensitivities for nodules equal to or less than 5 mm in diameter were reportedly less than 45%, although various sequences such as electrocardiograph (ECG)-triggered proton density weighted (PDW) or T2-weighted turbo spin echo (SE), ECG-triggered PDW black-blood-prepared half-Fourier single-shot turbo SE (HASTE), respiratory-triggered T2-weighted short-inversion-time inversion recovery (STIR) turbo SE, pre- and post-contrast-enhanced Volumetric interpolated breath-hold examination (VIBE), T2-weighted triple inversion black blood turbo spin echo, etc. were tested with nodules equal to or more than 5 mm section thickness on 1.5 or 3.0 T scanners.

The detection rates or sensitivities of MR imaging using various sequences on 1.5 and 3.0 T systems have ranged from 36.0 to 96.0% (Kersjes et al. 1997; Vogt et al. 2004; Schroeder et al. 2005; Luboldt et al. 2006; Bruegel et al. 2007; Regier et al. 2007; Yi et al. 2007; Frericks et al. 2008; Koyama et al. 2008; Sommer et al. 2014). The study with 150 subjects considering as the large population to date compared CT and MR imaging on a 1.5 T system demonstrated that the overall detection rate of thin-section CT (97%) was superior to that of respiratory-triggered STIR turbo SE imaging (82.5%), although there were no significant differences between the methods in the detection rate for all types of malignant nodules (Koyama et al. 2008). Therefore, the currently available MR technique should be considered as capable of performing a complimentary function for intrathoracic metastasis detection, and radiologists need to carefully check the nodules on thoracic MR images of their oncology patients.

In addition, more recent study demonstrated that pulmonary thin-section MR imaging with ultrashort echo time (UTE) had almost similar capability for nodule detection as compared with standard- and low-dose thin-section CT in routine clinical practice (Ohno et al. 2016). Therefore, further improvements of MR systems and sequences can be expected to enable pulmonary metastasis surveillance and/or lung cancer screening as well as morphological lung nodule characterization similar to CT on not only dedicated chest but also whole-body MR imaging in the near future (Fig. 1).

Fig. 1
figure 1figure 1

A 71-year-old female with adenomatous hyperplasia in the right lower lobe. Thin-section CT indicates ground-glass attenuation in the right lower lobe (a). Black-blood T1-weighted turbo SE (b), T2-weighted turbo SE (c), and STIR turbo SE (d) images show a nodule as, respectively, low, intermediate, and high signal intensity in the right lower lobe

Full size image

3 Characterization and Management of Pulmonary Nodules or Masses on MR Imaging

Since the pulmonary nodule is one of the most common findings on chest radiographs and CT, it is important to differentiate malignant from benign nodules in the least invasive way and to make as specific and accurate a diagnosis as possible. Investigators have used CT, MR imaging, and FDG-PET or PET/CT to evaluate the radiological features, relaxation time, blood supply, and metabolism of pulmonary nodules to differentiate malignant from benign nodules with promising results (Table 1).

Table 1 Recommended MR protocol for nodule detection and characterization
Full size table

3.1 Conventional T1-Weighted and T2-Weighted MRI Without and with Contrast Media

Characterization of the primary tumor on CT and MR imaging is based on the imaging features of the nodule or mass itself and its relationship to the pleura, chest wall, airways, and mediastinum, as well as its relative enhancement by contrast media. Historically, non-contrast-enhanced MR imaging has shown limited potential for characterizing peripheral lung nodules and masses and identifying benign nature of nodules due to the low intrinsic signal intensity of the lung parenchyma, the relatively poor spatial resolution, and patient-related motion artifacts (Caskey et al. 1990; Feuerstein et al. 1992; Kono et al. 1993; Kersjes et al. 1997; McLoud and Swenson 1999). In general, many pulmonary nodules, including lung cancers, pulmonary metastases and low-grade malignancies such as carcinoids and lymphomas are demonstrated as low or intermediate signal intensities on T1-weighted images and as slightly high intensity on T2-weighted images when SE or turbo SE sequences are used (Caskey et al. 1990; Feuerstein et al. 1992; Kono et al. 1993; Kersjes et al. 1997; McLoud and Swensen 1999) (Fig. 2). Malignant pulmonary nodules less than 30 mm in diameter usually do not show macroscopic necrosis (Caskey et al. 1990; Feuerstein et al. 1992; Kono et al. 1993; Kersjes et al. 1997; McLoud and Swensen 1999). Although enhancement levels vary due to underlying microscopically determined pathologic conditions such as tumor angiogenesis, tumor interstitial spaces, the presence or absence of fibrosis, and scarring and necrosis within the tumor, malignant pulmonary nodules show homogeneous enhancement but at a variety of levels on T1-weighted images after administration of contrast media (Caskey et al. 1990; Feuerstein et al. 1992; Kono et al. 1993; Kersjes et al. 1997; McLoud and Swensen 1999). Consequently, when using pre- and post-contrast-enhanced conventional T1-weighted images and T2-weighted images, clinicians in routine clinical practice often face a diagnostic dilemma in distinguishing malignant from benign pulmonary nodules such as organizing pneumonia, benign tumors, and inflammatory nodules (Caskey et al. 1990; Feuerstein et al. 1992; Kono et al. 1993; Kersjes et al. 1997; McLoud and Swensen 1999) (Figs. 2 and 3). It has therefore been suggested that enhancement patterns or blood supply evaluated with dynamic contrast-enhanced MR imaging may be helpful for diagnosis and management of pulmonary nodules (Kono et al. 1993, 2007; Kusumoto et al. 1994; Hittmair et al. 1995; Gückel et al. 1996; Fujimoto et al. 2003; Ohno et al. 2002, 2004a; Schaefer et al. 2004, 2006; Donmez et al. 2007). However, it may be possible to diagnose several histological types of pulmonary nodules, such as bronchocele, tuberculoma, mucinous bronchoalveolar carcinoma (BAC), hamartoma, and aspergilloma, on pre- or post-contrast-enhanced T1-weighted images and T2-weighted image according to their specific MR findings.

Fig. 2
figure 2figure 2

A 56-year-old female with adenocarcinoma in the right lower lobe. Thin-section CT shows a solid nodule in the right lower lobe (a). Black-blood T1-weighted turbo SE (b), T2-weighted turbo SE (c), and STIR turbo SE (d) images show nodule as, respectively, low, intermediate, and high signal intensity in the right lower lobe. Post-contrast black-blood T1-weighted turbo SE image (e) showed homogeneous enhancement of the nodule

Full size image
Fig. 3
figure 3figure 3

A 55-year-old male with cryptococcosis in the left lower lobe. Thin-section CT shows a solid nodule in the left lower lobe (a). Black-blood T1-weighted turbo SE (b), T2-weighted turbo SE (c), and STIR turbo SE (d) images show nodule as, respectively, low, intermediate, and high signal intensity in the left lower lobe. Post-contrast black-blood T1-weighted turbo SE image (e) showed homogeneous enhancement of nodule

Full size image

3.1.1 Bronchocele

Bronchial atresia is a common focal pulmonary lesion, which can be diagnosed by using non-contrast-enhanced T1- and T2-weighted images. Bronchial atresia is an uncommon anomaly characterized by focal obliteration of the bronchial lumen and the absence of communication between lobar, segmental, or subsegmental bronchi and the central airway (Meng et al. 1978; Jederlinic et al. 1987; Finck and Milne 1988; Naidich et al. 1988; Bailey et al. 1990; Ko et al. 1998; Matsushima et al. 2002). Mucus secreted within the patent airways distal to the point of atresia accumulates in the form of a plug or bronchocele which appears as a pulmonary nodule or mass (Finck and Milne 1988; Naidich et al. 1988; Bailey et al. 1990; Ko et al. 1998; Matsushima et al. 2002). The MR image of bronchoceles reportedly appears as a branching lesion with high signal intensity on T1- and T2-weighted images due to the dilated mucus-filled bronchi and mucocele formation distal to the atelectatic segment (Finck and Milne 1988; Naidich et al. 1988; Bailey et al. 1990; Ko et al. 1998; Matsushima et al. 2002) (Fig. 4).

Fig. 4
figure 4figure 4

A 15-year-old female with bronchocele in the right upper lobe. Chest radiograph shows a mass in the right upper lung field (a). Contrast-enhanced CT shows a mass with water density in the right upper lobe (b). Conventional T1-weighted SE (c) and T2-weighted turbo SE (d) images show the mass as high intensity due to mucus-filled bronchi and formation of a mucocele distal to the atelectatic segment, which was diagnosed as a bronchocele

Full size image

3.1.2 Tuberculoma

Tuberculomas, observed as well-defined nodules located mainly in the upper lobes, may appear after primary or reactivated tuberculosis. Calcification occurs in about 20–30% of cases (Sochocky 1958). CT yields superior visualization of the calcifications and characteristics of the nodules. In rare cases, areas of diminished attenuation are seen, which represent caseous necrosis. The latter may be identified as tuberculoma, a nodule with a relatively low signal intensity in comparison with that of other pulmonary nodules on T2-weighted SE or turbo SE images (Sakai et al. 1992; Kono et al. 1993; Kusumoto et al. 1994; Parmar et al. 2000; Chung et al. 2000; Schaefer et al. 2006). In addition, several investigators have reported on typical MR findings of tuberculoma on post-contrast-enhanced T1-weighted images known as “thin-rim enhancement” sign (Sakai et al. 1992) (Fig. 5). Signal intensity at the center of tuberculomas is low or slightly enhanced, but the signal intensity of the fibrotic rim is markedly enhanced. These MR findings correspond well to those of pathological specimens (Sakai et al. 1992; Kono et al. 1993; Kusumoto et al. 1994; Parmar et al. 2000; Chung et al. 2000; Schaefer et al. 2006). Therefore, when tuberculoma is suspected or attempts are made to distinguish it noninvasively from malignant nodules, pre- and post-contrast-enhanced conventional MR imaging may be the most suitable procedure because tuberculoma is one of the most well-known diseases that show intense FDG uptake and is difficult to distinguish from malignant nodules when FDG-PET or PET/CT is used (Chang et al. 2006).

Fig. 5
figure 5figure 5

A 67-year-old male with tuberculoma in the left lower lobe. Thin-section CT shows a solid nodule in the left lower lobe (a). Contrast-enhanced CT indicates homogeneity but little enhancement within the nodule (b). Black-blood T1-weighted turbo SE (c) and T2-weighted turbo SE (d) images show the nodule as, respectively, low and intermediate signal intensities in the left lower lobe. Post-contrast black-blood T1-weighted turbo SE image (e) showed thin-rim enhancement, which was diagnosed as tuberculoma. Dynamic MR imaging (f) with ultrafast GRE technique (L to R: t = 0.0 s, t = 8.8 s, t = 16.5 s, and t = 25.4 s) showed a slightly enhancing nodule in the left lower lung field. The nodule is not enhanced in the lung parenchymal phase (t = 8.8 s) and slightly enhanced in the systemic circulation phase (t = 16.5 and 25.4 s) due to low blood supply from bronchial circulation

Full size image

3.1.3 Mucinous Bronchoalveolar Carcinoma

Adenocarcinoma of the lung constitutes a histologically and biologically heterogeneous group of tumors. Except for mucinous BAC, mucin production is seldom a truly prominent characteristic of this adenocarcinoma, so that there are no significant differences in the MR findings for adenocarcinoma of the lung and other subtypes of lung cancer. Mucinous BAC occasionally presents as a solitary pulmonary nodule with low or no uptake of contrast media on conventional post-contrast-enhanced T1-weighted image and thus may be difficult to differentiate from other benign nodules such as tuberculoma. However, many mucinous BACs demonstrate one of two radiological patterns, consolidation or diffuse disease. The consolidation may be segmental or may involve an entire lobe and, with the exception of pulmonary vasculature, shows as high signal intensity on T2-weighted turbo SE images, which is known as “white lung sign” (Fig. 6). In addition, intratumoral vessels may be detected on contrast-enhanced T1-weighted images (Gaeta et al. 2000, 2001, 2002).

Fig. 6
figure 6figure 6

A 73-year-old male with mucinous bronchoalveolar carcinoma (BAC) in the right middle and both lower lobes. Conventional CT shows ground-glass attenuation in the right middle and both lower lobes (a). Black-blood T1-weighted turbo SE (b) image shows mucinous BAC as low signal intensities. Black-blood T2-weighted turbo SE (c) image shows mucinous BAC with high signal intensities. These findings are known as “white lung sign”

Full size image

3.1.4 Hamartoma

Pulmonary hamartomas, the third most common cause of solitary pulmonary nodules, are considered benign neoplasms that originate in fibrous connective tissue beneath the mucus membrane of the bronchial wall (Bateson 1973; Siegelman et al. 1986). A few investigators have reported on MR imaging of hamartomas. Sakai et al. (1994) found that all hamartomas appeared as a signal of intermediate intensity on T1-weighted SE images and as one of high intensity on T2-weighted SE images and four out of six hamartomas had a lobulated appearance separated by septa on T1- or T2-weighted SE images. In addition, pulmonary hamartomas pathologically diagnosed as lipomatous hamartoma sometimes show with high signal intensity on T1-weighted SE and T2-weighted turbo SE images (Yilmaz et al. 2004). Therefore, when these radiological findings are associated with a pulmonary nodule, the nodule can be suspected of being a hamartoma, and the enhancement pattern within the nodule should be evaluated on post-contrast-enhanced T1-weighted SE images, where the regions with less enhancement correspond to core cartilaginous tissue and septa and areas of marked contrast enhancement correspond to branching cleft-like mesenchymal connective tissue that dip into the cartilaginous core (Sakai et al. 1994) (Fig. 7).

Fig. 7
figure 7figure 7

A 54-year-old male with hamartoma in the left upper lobe. Conventional contrast-enhanced CT shows lobulated nodule in the left upper lobe (a). Black-blood T1-weighted turbo SE (b) and T2-weighted (c) images show a lobulated nodule as, respectively, low and high signal intensities in the left upper lobe. Post-contrast black-blood T1-weighted turbo SE (d) image shows less enhancement corresponding to core cartilaginous tissue, while areas of marked contrast enhancement correspond to cleft-like branching mesenchymal connective tissue

Full size image

3.1.5 Aspergilloma

Saprophytic aspergillosis (aspergilloma) is characterized by Aspergillus infection without tissue invasion. It typically leads to conglomeration of intertwined fungal hyphae admixed with mucus and cellular debris within a pre-existing pulmonary cavity or ectatic bronchus (Gefter 1992; Aquino et al. 1994). A typical radiological finding of aspergilloma is a solid, round, or oval mass with soft-tissue opacity within a lung cavity. The mass is typically separated from the cavity wall by an airspace manifesting the so-called “air crescent” sign and is often associated with thickening of the wall and adjacent pleura (Gefter 1992; Park et al. 2007). The most common underlying causes are healed tuberculosis, bronchiectasis, bronchial cyst, and sarcoidosis. Other conditions that occasionally may be associated with aspergilloma include bronchogenic cyst, pulmonary sequestration, and pneumatoceles secondary to Pneumocystis carinii pneumonia in patients with acquired immunodeficiency syndrome.

The presence of a cavity within lung cancer is common and has been reported in 2–16% of cases (Felson and Wiot 1977). In general, lung cancer with a cavity typically shows a cavity wall with a thickness of more than 4 mm, spiculate or irregular inner and outer margins, enlarged lymph nodes, and a soft-tissue nodule or mass due to intracavitary tumorous mural regions associated with infiltration of the adjacent thoracic wall. In some cases, however, a lung cancer with a cavity may show a thin wall or “air crescent” sign on CXR or CT. It is therefore important to distinguish fungus ball from intracavitary tumorous mural regions by the “air crescent” sign in lung cancers with cavities, and pre- and post-contrast-enhanced conventional T1- and T2-weighted SE or turbo SE images are helpful for this purpose.

Although a cavity may be present within lung cancer, a viable lung cancer evidenced as a cavity and/or intracavitary mass shows a typical signal intensity pattern on pre- and post-contrast-enhanced conventional T1- and T2-weighted SE or turbo SE images with as its characteristics a signal of low or intermediate intensity on pre-contrast-enhanced T1-weighted images, one of intermediate or high intensity on T2-weighted images, and one of high intensity due to intensive and homogeneous enhancement on post-contrast-enhanced T1-weighted images (Ma et al. 1985; van der Heide et al. 1985; Zinreich et al. 1988; Herold et al. 1989; Fujimoto et al. 1994; Blum et al. 1994) (Fig. 8). In aspergilloma cases, on the other hand, the signal intensity of the intracavitary lesion is especially reduced on the T2-weighted SE or turbo SE image because of the presence of calcium, air, or ferromagnetic elements resulting from Aspergillus infection (Herold et al. 1989; Fujimoto et al. 1994; Blum et al. 1994) (Fig. 9). Because the presence of the ferromagnetic elements of iron, magnesium, and manganese is essential to the metabolism of amino acids by fungi, Fujimoto et al. (1994) have suggested that the reduction in signal intensity on T1-weighted images and the marked reduction on T2-weighted images are characteristics of aspergilloma as well as mycetomas and may be useful for differentiation of aspergilloma from intracavitary tumorous mural nodules.

Fig. 8
figure 8figure 8

A 74-year-old male with squamous cell carcinoma in the left lower lobe. Thin-section CT (a) and sagittal reformat (b) show mass with irregularly thickened cavity wall in the left lower lobe. Black-blood T1-weighted turbo SE (c) and T2-weighted (d) images show mass with irregularly thickened cavity wall as, respectively, low and high signal intensities in the left lower lobe. Post-contrast black-blood T1-weighted turbo SE (e) image shows well-enhanced cavity wall corresponding to lung cancer

Full size image
Fig. 9
figure 9figure 9

A 60-year-old male with aspergilloma in the left upper lobe. Chest radiograph (a) shows “air crescent” sign in the left upper lung field. Conventional CT (b) shows fungus ball within cavity. Conventional T1-weighted SE (c) and T2-weighted turbo SE (d) images on coronal plane show intracavitary lesion as, respectively, low and very low signal intensities in the left upper lobe. Post-contrast T1-weighted SE image (e) shows well-enhanced cavity wall and no enhancement of intracavitary lesion

Full size image

3.2 New Non-Contrast-Enhanced MR Imaging of Pulmonary Nodules

To overcome limitations in differentiating benign from malignant nodules, STIR turbo SE imaging and diffusion-weighted imaging (DWI) were introduced in 2008 as more promising sequences than T2WI and pre- or post-CE T1WI for non-CE MR imaging for nodule assessment (Koyama et al. 2008; Mori et al. 2008). Koyama et al. demonstrated that the quantitative capability of STIR turbo SE imaging was significantly better than non-CE T1WI or T2WI for differentiating malignant from benign nodules, with sensitivity, specificity, and accuracy of 83.3%, 60.6%, and 74.5%, respectively (Koyama et al. 2008). DWI is usually assessed by the apparent diffusion coefficient (ADC), which evaluates the diffusivity of water molecules within tissue between 0 s/mm2 and a maximum b value ranging from 500 to 1000 s/mm2 in routine clinical practice. According to this ADC evaluation, the quantitative and/or qualitative sensitivities and specificities of DWI in this setting have been reported as 70.0–88.9% for sensitivity and 61.1–97.0% for specificity (Mori et al. 2008; Satoh et al. 2008; Uto et al. 2009). On the other hand, one study has suggested the signal intensity ratio between lesion and spinal cord is more useful than ADC and used this ratio to determine that sensitivity, specificity, and accuracy of DWI were 83.3%, 90.0%, and 85.7%, respectively, and that the accuracy of this new parameter was significantly higher than that of ADC (50.0%) (Uto et al. 2009). It would therefore be better to use the abovementioned two non-CE MR techniques as routine clinical protocols non-CE-T1WI, T2WI, and CE-T1WI to improve differentiation capability between malignant and in addition to benign nodules on non-CE-MR examinations.

3.3 Dynamic Contrast-Enhanced MR Imaging of Pulmonary Nodules

Although in some cases benign or malignant focal lesions can be differentiated from others by using pre- and post-contrast-enhanced conventional T1- or T2-weighted SE or turbo SE imaging, significant overlaps have been observed between benign and malignant lesions in routine clinical practice (Kono et al. 1993; Kusumoto et al. 1994). To overcome this problem, dynamic contrast-enhanced MR imaging has been proposed as an alternative technique for diagnosis and/or management of pulmonary nodules (Table 1). As a result of advances in MR systems and pulse sequences, there are now three major methods for dynamic MR imaging of the lung. Many investigators have proposed dynamic MR imaging be used for two-dimensional (2D) SE or turbo SE sequences or for various types of 2D or three-dimensional (3D) gradient-echo (GRE) sequences and that enhancement patterns within nodules and/or parameters determined from signal intensity-time course curves be assessed visually. These curves represent first transit and/or recirculation and washout of contrast media in 5 min or more with repeated breath holding (Kono et al. 1993, 2007; Kusumoto et al. 1994; Hittmair et al. 1995; Gückel et al. 1996; Ohno et al. 2002, 2004a; Fujimoto et al. 2003; Schaefer et al. 2004, 2006; Donmez et al. 2007). Taking into account the inherent inhomogeneous composition of many intraindividual lung cancers and even within benign lesions such as central necrosis, only a dynamic contrast-enhanced 3D approach has the chance of depicting underlying histologies. This is important for the primary diagnosis as well as during follow up examinations to assure evaluation of the same region of interest again. Recently, dynamic contrast-enhanced MR perfusion imaging has been proposed for quantitative and qualitative assessment of regional pulmonary perfusion abnormalities by using 2D or 3D ultrafast GRE sequences with sharp bolus profiles (Hatabu et al. 1996, 1999; Amundsen et al. 1997, 2000; Levin et al. 2001; Matsuoka et al. 2001; Ohno et al. 2002, 2004c, 2005, 2007a, b; Fink et al. 2004). This technique allows for directly assessing the first passage of contrast media within nodules in less than 35 s within a single breath hold and to evaluate blood supply to nodules from pulmonary and/or bronchial circulation (Ohno et al. 2002, 2004c). It should be noted though that the ideal acquisition of contrast-enhanced dynamic studies of tumor perfusion should not be limited to the first pulmonary passage of the contrast agent. Tumors will be supplied by systemic, bronchial arteries instead of pulmonary arteries (Milne and Zerhouni 1987; Ohno et al. 2002, 2004c). In this case or due to inherent slow blood flow within a tumor an examination in a single breath hold might not be sufficient in detecting the full perfusion cycle of wash-in and wash-out (Ohno et al. 2002, 2004c). Examination of prolonged wash-out always reveals an underlying reperfusion of pulmonary tissue by both pulmonary and bronchial arteries. This circumstance has not been accounted for in the usual models for quantification of tissue perfusion so far.

Although there are various dynamic MR techniques, reported sensitivities range from 94 to 100%, specificities from 70 to 96%, and accuracies of more than 94% (Kono et al. 1993, 2007; Patz et al. 1993; Kusumoto et al. 1994; Hittmair et al. 1995; Gückel et al. 1996; Ohno et al. 2002, 2004a; Fujimoto et al. 2003; Schaefer et al. 2004, 2006; Donmez et al. 2007). These specificities and accuracies for dynamic MR imaging are superior to those reported for dynamic CT and almost equal to or superior to those for FDG-PET or PET/CT (Swensen et al. 1992, 1996, 2000; Dewan et al. 1993; Kono et al. 1993; Kusumoto et al. 1994; Hittmair et al. 1995; Yamashita et al. 1995; Bury et al. 1996; Gückel et al. 1996; Zhang and Kono 1997; Ohno et al. 2002, 2004a; Fujimoto et al. 2003; Herder et al. 2004; Schaefer et al. 2004, 2006; Yi et al. 2004, 2006; Jeong et al. 2005; Joshi et al. 2005; Mori et al. 2005; Bryant and Cerfolio 2006; Christensen et al. 2006; Donmez et al. 2007; Kim et al. 2007; Kono et al. 2007; Lee et al. 2007). Therefore, dynamic contrast-enhanced MR imaging may play a complementary role in the diagnosis of pulmonary nodules assessed with FDG-PET or PET/CT. Although these results are highly promising, further research in this field may be necessary in the light of recently published data regarding the MR signal dependency in perfusion studies on the contrast agent concentration. As quantification of perfusion parameters is dependent on an almost linear relationship of signal to concentration great care has to be taken regarding the dosage and application of contrast agents (Puderbach et al. 2008).

Although differentiation of malignant from benign pulmonary nodules by means of dynamic MDCT, dynamic MRI, and PET or PET/CT with FDG has been tried in several studies, accurate separation of active infectious nodules from malignant neoplasms on the basis of dynamic CT and MR parameters and uptakes of FDG can be extremely difficult in view of the underlying pharmacokinetical, pathological, and biological properties of malignant neoplasms and active infectious nodules. In addition, when considering the management of pulmonary nodules in clinical practice, it may be helpful to differentiate pulmonary nodules requiring further intervention and treatment (low- or high-grade malignant tumors and active infectious nodules) from pulmonary nodules requiring no further evaluation (benign tumors and chronic infectious nodules) rather than to differentiate between malignant nodules and other nodules. For this latter differentiation, ultrafast dynamic MR imaging can divide all nodules into the two categories (Figs. 5 and 10) (Ohno et al. 2002).

Fig. 10
figure 10figure 10

A 57-year-old female with invasive adenocarcinoma. Thin-section CT (a) and coronal reformat (b) show partly solid nodule in the right upper lobe. Dynamic MR imaging (c) with ultrafast GRE technique (L to R: t = 0.0 s, t = 8.8 s, t = 11.0 s, and t = 25.4 s) shows well-enhanced nodule (arrow) in the right upper lung field. The nodule was not enhanced in the lung parenchymal phase (t = 8.8 s) and well enhanced in the systemic circulation phase (t = 11.0 and 25.4 s) due to enhanced blood supply from bronchial circulation

Full size image

In addition, the results of several comparative studies of findings obtained with dynamic MR parameters and with immuno-histopathological examination of small peripheral lung cancers, the former showed good correlation with tumor angiogenesis (Fujimoto et al. 2003; Schaefer et al. 2006), and the potential for more accurate differentiation of subtypes of small peripheral adenocarcinomas than is possible with thin-section CT, or for prognosis both before and after treatment (Fujimoto et al. 2003; Ohno et al. 2005; Schaefer et al. 2006).

In the last a few decades, the diagnostic performance of these dynamic MR techniques for distinguishing malignant from benign nodules has been reported as comprising sensitivity ranging from 94 to 100%, specificity from 70 to 96%, and accuracy of more than 94% (Gückel et al. 1996; Ohno et al. 2002, 2004a, 2008a, 2014; Fujimoto et al. 2003; Schaefer et al. 2004; Kono et al. 2007; Cronin et al. 2008). In addition, a meta-analysis found that there were no significant differences in diagnostic performance among dynamic CE-CT, dynamic CE-MR imaging, FDG-PET, and single photon emission tomography (SPECT) (Cronin et al. 2008), although dynamic MR imaging with the 3D GRE sequence and ultrashort TE proved its superior diagnostic performance in a direct and prospective comparison study of dynamic CE-CT and co-registered FDG-PET/CT (Ohno et al. 2008a). The use of dynamic MR imaging with the 3D GRE sequence and ultrashort TE is therefore likely to be more effective than that of other methods and may lead to improved diagnostic performance of dynamic CE-MR imaging. In addition, this method has the potential to play a complementary or substitutional role in the characterization of pulmonary nodules assessed with dynamic CE-CT, FDG-PET, and/or PET/CT.

Moreover, ultrafast 3D dynamic MR data can also be used for the prediction of postoperative lung function for NSCLC patients (Ohno et al. 2004b, 2007b, 2011a, b). The semiquantitative regional perfusion obtained from ultrafast dynamic MR imaging shows good correlation with that assessed by perfusion scintigraphy, with a reported limit of agreement of ± 6%, which is insignificant enough to make ultrafast dynamic MR imaging suitable for clinical purposes (Ohno et al. 2004b). Moreover, the prediction of postoperative lung function derived from dynamic MR imaging was more accurate than that derived from perfusion scintigraphy, single photon emission tomography (SPECT), SPECT fused with CT (SPECT/CT), or quantitatively and/or qualitatively assessed MDCT, which is predicted postoperative lung function from preoperative lung function and the number of lung segments in the total and resected lung evaluated by pulmonary surgeons, and the predictive accuracy is almost equal to that obtained with quantitatively assessed MDCT based on density-masked CT technique (Ohno et al. 2004b, 2007b, 2011a). Dynamic MR imaging may therefore be useful not only for the characterization of pulmonary nodules, but prediction of postoperative lung function may also assist the management of pulmonary nodules, including determination of whether further intervention and treatment and/or operability are indicated for lung cancer patients.

4 Assessment of TNM Stages

The international TNM classification proposed by the International Union against Cancer (UICC) has been widely used in the investigation and treatment of lung cancer (Tables 2, 3, 4, and 5), and shows that survival rates have improved with more accurate staging and more accurate differentiation between those patients who are candidates for surgical resection and those who are judged to be inoperable but would benefit from chemotherapy, radiotherapy, or both (Sobin and Wittekind 2002). Therefore, accurate radiological staging may affect the management as well as the prognosis of patients. However, only approximately one-half of the TNM stages determined with CT systems in the past have agreed with operative staging, with patients being both under- and over-staged (Lewis et al. 1990; Gdeedo et al. 1997).

Table 2 Stage according to Goldstraw et al. (2007)
Full size table
Table 3 T classification – primary tumor according to Goldstraw et al. (2007)
Full size table
Table 4 N classification – regional lymph nodes according Goldstraw et al. (2007)
Full size table
Table 5 M classification – distant metastasis according Goldstraw et al. (2007)
Full size table

Currently, newly developed MDCT systems, FDG-PET or PET/CT, are considered useful for precise assessment of tumor extent because of their multiplanar capability and for accurate diagnosis of metastatic lymph nodes by analyzing the glucose metabolism of cancer cells in lung cancer patients. However, since 1991 it has been suggested that MR imaging, with its multiplanar capability and better contrast resolution of tumor and mediastinum or of tumor and chest wall or both than that of CT, may also be useful, but only for the assessment of mediastinal and chest wall invasions and determination of the short axis diameter of certain mediastinal lymph nodes (Webb et al. 1991). However, recent advancements in MR systems, improved or newly developed pulse sequences and/or utilization of contrast media has resulted in improved diagnosis of TNM staging for lung cancer patients (Table 6).

Table 6 Recommended MR protocol for lung cancer staging
Full size table

4.1 MR Assessment of T Classification

T classification is the descriptor given to the primary tumor and its local extent (Sobin and Wittekind 2002). The definitions are given in Table 3. While the T factor is subdivided into four groups, the distinction between T3 and T4 tumors is critical because it represents the dividing line between conventional surgical and nonsurgical management for NSCLC patients (Armstrong 2000). It is therefore more important to distinguish T3 from T4 tumors than to differentiate T1 and T2 tumors and determine nodal staging. For this reason, MR imaging may be helpful for assessment of mediastinal invasion, chest wall invasion, and distinguishing primary tumors from secondary changes such as atelectasis or obstructive pneumonia, although it may be difficult to distinguish simple extension of the tumor into the mediastinal pleura or pericardium (T3) from actual invasion (T4).

4.1.1 Mediastinal Invasion

Many surgeons consider minimal invasion of mediastinal fat as resectable (Quint and Francis 1999), so that clinicians want to know whether minimal mediastinal invasion (T3 disease) or actual invasion (T4 disease) has occurred before considering surgical resection. The accuracy of CT for evaluating hilar and mediastinal invasion of lung cancer has been investigated extensively over the last decades. Sensitivity for assessment of mediastinal invasion by single detector computed tomography with or without the use of helical scanning ranged from 40 to 84% and specificity from 57 to 94% (Baron et al. 1982; Martini et al. 1985; Quint et al. 1987, 1995; Rendina et al. 1987; Glazer et al. 1989; Herman et al. 1994; White et al. 1994; Takahashi et al. 1997).

The RDOG study compared CT with MR imaging for 170 patients with NSCLC, although only T1-weighted images obtained without the use of cardiac or respiratory gating techniques were assessed in this study (Webb et al. 1991). Although there was no significant difference between the sensitivity (63% and 56% for CT and MR imaging, respectively) and the specificity (84% and 80%) for distinguishing between T3-T4 tumors and T1-T2 tumors in this study, the RDOG reported that 11 patients showed mediastinal invasion and that the superior contrast resolution of MR imaging conferred a slight but statistically significant advantage over CT for accurate diagnosis of mediastinal invasion. In addition, delineation of tumor invasion of pericardium (T3) or heart (T4) was superior on MR imaging compared with CT scan when the cardiac-gated T1-weighted sequence was used for improved avoidance of cardiac motion artifacts (White 1996). The normal pericardium has low signal intensity. Direct invasion of the cardiac chambers is readily demonstrated on T1-weighted images, because blood flowing through the cardiac chambers is signal void, so that the tumor is conspicuous because of its higher signal intensity. However, the accuracy of minimal mediastinal invasion assessment by both CT scanning and MR imaging is limited because it depends on visualization of the tumor within the mediastinal fat (Wong et al. 1999). In contrast to the assessment of mediastinal invasion, Mayr et al. (1987) found CT scanning to be more accurate than MR imaging in visualizing and assessing both normal and abnormal airways. They evaluated 319 normal and 79 abnormal bronchoscopically visualized bronchi. Their study found that CT scanning was accurate in all cases, whereas MR imaging correctly identified only 45% of normal bronchi and 72% of abnormal bronchi (Magdeleinat et al. 2001). This discrepancy can be attributed both to the higher spatial resolution of CT scanning and to the low intrinsic MR imaging signal of air. Therefore, the relationship of lung cancer to central endobronchial extension is more accurately demonstrated on CT scans.

Recent advancement in MR systems, improved pulse sequences, and utilization of contrast media have resulted in the introduction of new MR imaging techniques for assessment of mediastinal invasion of lung cancer. Contrast-enhanced MR angiography has been used for assessment of cardiovascular or mediastinal invasions (Takahashi et al. 2000; Ohno et al. 2001). Ohno et al. (2001) described a series of 50 NSCLC patients with suspected mediastinal and hilar invasion of lung cancer visualized with contrast-enhanced CT scans, cardiac-gated MR imaging, and non-cardiac- and cardiac-gated contrast-enhanced MR angiographies (Fig. 11). In this study, sensitivity, specificity, and accuracy of contrast-enhanced MR angiography ranged from 78% to 90%, 73% to 87%, and 75% to 88%, respectively. These values were higher than those of contrast-enhanced single helical CT and conventional T1-weighted imaging (Ohno et al. 2001). Thus, contrast-enhanced MR angiography is thought to improve the diagnostic capability of MR imaging for mediastinal and hilar invasion.

Fig. 11
figure 11figure 11

A 68-year-old female with squamous cell carcinoma in the left upper lobe. Contrast-enhanced thin-section CT (a) suggests invasion to pulmonary vein and mediastinum, while conventional T1-weighted image (b) on coronal plane suggests invasion to pulmonary artery and mediastinum. Contrast-enhanced MR angiography (c) clearly shows invasion to pulmonary artery and vein, but not mediastinal invasion due to visualization of mediastinal fat between pulmonary artery and pulmonary vein as a black area. Macroscopical finding of the resected left upper lobe (d) shows invasion to left pulmonary artery (arrow) and left superior pulmonary vein (large arrow), but not mediastinal invasion

Full size image

In 2005, another new technique, cine MR imaging obtained with a steady-state free precession (SSFP) sequence was introduced as useful for evaluation of cardiovascular invasion in patients with thoracic mass (Seo et al. 2005). In this study, as well as previous electron beam CT or traditional cine MR studies (Murata et al. 1994; Sakai et al. 1997), the assessment of sliding motions between thoracic masses and adjacent mediastinal structure demonstrated a very high diagnostic capability (sensitivity: 100%, specificity: 92.9%, accuracy: 94.4%) (Seo et al. 2005). However, only 9 of 26 lung cancer patients were included in this study since the others had mediastinal tumors. Further investigation thus seems to be warranted to determine the actual diagnostic capability of cine MR imaging for mediastinal invasion in NSCLC patients.

MDCT is widely utilized for routine clinical practice, and it was found that thin-section multiplanar reformatted (MPR) imaging from thin-section volumetric MDCT data was useful for the evaluation of T classification due to its multiplanar capability (Higashino et al. 2005). Higashino et al. (2005) suggested that mediastinal invasion that can be assessed from thin-section coronal MPR images with 1 mm section thickness with greater sensitivity, specificity, and accuracy than can be achieved with routine MDCT with 5 mm section thickness and with slightly better specificity and accuracy than with thin-section axial MDCT with 1 mm section thickness. Although MR imaging is considered to show superior tissue contrast to that of MDCT, the similar multiplanar capability, faster scan time and better spatial resolution of thin-section MDCT may result in better assessment of mediastinal invasion in NSCLC patients than by previously described MR techniques (Fig. 12). Further investigations as well as comparative studies of thin-section MDCT and previously described or newly developed MR imaging techniques thus seem to be warranted to determine the actual significance of MR imaging for assessment of mediastinal invasion in routine clinical practice.

Fig. 12
figure 12figure 12

A 61-year-old male with squamous cell carcinoma in the right upper lobe. Contrast-enhanced CT (a) suggests mediastinal invasion, while contrast-enhanced coronal reformat (b) clearly demonstrates mediastinal invasion. Black-blood T1-weighted images on axial (c) and coronal (d) planes also clearly demonstrate invasion to superior venous cava and mediastinum

Full size image

4.1.2 Distinguishing Lung Cancer from Secondary Change

Distinguishing primary lung cancer from secondary change is important for assessment of tumor extent and the therapeutic effect of chemotherapy and/or radiotherapy. While the therapeutic effect of conservative therapy has been assessed by using World Health Organization (WHO) criteria or response evaluation criteria in solid tumors (RECIST) (World Health Organization 1979; Therasse et al. 2000), it would be difficult to use CXR or plain or contrast-enhanced CT to evaluate tumor extent or therapeutic effect for cases with atelectasis or obstructive pneumonia.

MR imaging, on the other hand, has potential for distinguishing lung cancer from secondary change due to atelectasis or pneumonitis (Kono et al. 1993). In some cases, it can be difficult to distinguish lung cancer from post-obstructive atelectasis or pneumonitis because these secondary changes tend to be enhanced to a similar degree as the central tumor on contrast-enhanced CT scan. On T2-weighted MR imaging, however, post-obstructive atelectasis and pneumonitis often show higher signal intensity than does the central tumor. Bourgounin et al. (1991) evaluated the histological findings of obstructive pneumonitis or atelectasis in patients who had undergone surgical resection of lung cancer and had been evaluated preoperatively with MR imaging. They found that cholesterol pneumonitis and mucus plugs displayed higher signal intensity than the tumor on T2-weighted images, while atelectasis and organizing pneumonitis were isointense to the tumor. Kono et al. (1993) described a series of 27 patients with central lung cancer associated with atelectasis or obstructive pneumonitis (Fig. 13). These patients were examined with post-contrast-enhanced T1-weighted MR imaging, and the central tumor could be differentiated from adjacent lung parenchymal disease in 23 out of 27 patients (85%). The tumor was of lower signal intensity than the adjacent lung disease in 18 cases (67%) and of higher signal intensity in 5 (18%). These differences in signal intensity between primary tumor and secondary change were considered to be due to the presence of invasion of pulmonary vasculatures. Therefore, the use of T2-weighted or post-contrast-enhanced T1-weighted images for assessment of tumor size and secondary change may be helpful for precise evaluation of tumor extent at the initial staging and for accurate prognosis for patients and assessment of therapeutic effect after conservative therapy and/or for comparative studies of standard and new chemo- and/or radiotherapy regimens (Ohno et al. 2000).

Fig. 13
figure 13figure 13

A 59-year-old male with adenocarcinoma. Thin-section CT (a) shows atelectasis in the right middle and lower lobe. Contrast-enhanced thin-section CT (b) shows homogeneously enhanced atelectasis and pleural effusion, but the extent of the tumor cannot be determined. Black-blood T1-weighted (c) and T2-weighted (d) images show atelectasis and the tumor as, respectively, low and intermediate signal intensity. Mediastinal invasion is suspected, while pleural effusion is clearly seen. Post-contrast black-blood T1-weighted image (e) clearly shows lung cancer and secondary change (arrow)

Full size image

4.1.3 Chest Wall Invasion

Chest wall invasion used to be considered a contraindication for surgical excision of lung cancer, but recent surgical advances have made chest wall excision feasible for the treatment of locally aggressive lung cancer and giving patients a better chance of survival (Magdeleinat et al. 2001). Preoperative visualization of chest wall invasion may therefore be helpful for surgical planning. On conventional CT scan, rib destruction is the only reliable sign of chest wall invasion since soft-tissue masses in the chest wall correlate statistically with chest wall invasion. However, they are not reliable indicators for an individual patient, so that focal chest pain may still be the most reliable indicator of chest wall invasion. In fact, the reliability of conventional CT assessment of chest wall invasion in lung cancer patients varies widely with reported sensitivities ranging from 38 to 87% and specificities from 40 to 90% (Quint and Francis 1999). In addition to the technique of inducing artificial pneumothorax described elsewhere (Watanabe et al. 1991; Yokoi et al. 1991), Murata et al. (1994) reported that dynamic expiratory multi-section CT reviewed as a cine loop was 100% accurate for identification of both chest wall and mediastinal invasion.

Another study has suggested that ultrasound (US) is an effective technique for diagnosis of chest wall invasion (Suzuki et al. 1993). In this study, 120 lung cancer patients were examined, in 19 of whom invasion was pathologically proved. Sensitivity, specificity, and accuracy of US were 100%, 98%, and 98%, respectively, while the corresponding values for conventional CT used in the same study were only 68%, 66%, and 67% (Suzuki et al. 1993).

Because of its multiplanar capability and better tissue contrast resolution compared to CT, MR imaging has also been advocated as effective for assessment of chest wall invasion (Rapoport et al. 1988; Heelan et al. 1989; Webb et al. 1991; Padovani et al. 1993; Bonomo et al. 1996). Sagittal and coronal plane images are better than axial CT images for displaying the anatomical relationship between tumor and chest wall structures (Bonomo et al. 1996; Freundlich et al. 1996). MR imaging shows infiltration or disruption of the normal extrapleural fat plane on T1-weighted images or parietal pleural signal hyperintensity on T2-weighted images (Fig. 14). In addition, when STIR turbo SE imaging is used for this purpose, it can demonstrate lung cancer as high signal intensity within the suppressed signal intensities of chest wall structures, enabling clinicians to easily determine the tumor extent within chest wall (Fig. 14c). Moreover, Padovani et al. (1993) have suggested that the diagnostic yield can be further improved by intravenous administration of contrast media. In addition, superior sulcus or Pancoast tumors are good candidates for the demonstration of chest wall invasion on MR imaging. Since superior sulcus tumors occur in close proximity to the lung apex, their imaging has to include an evaluation of the relationship between the tumor and the brachial plexus, subclavian artery and vein, and adjacent chest wall. The axial scan plane of a CT scan is suboptimal for examining the lung apex where superior sulcus tumors are located, while direct sagittal and coronal MR images are superior to CT for evaluating the local extent of disease in patients with superior sulcus tumors (Rapoport et al. 1988; Heelan et al. 1989; Webb et al. 1991; Padovani et al. 1993; Bonomo et al. 1996; Freundlich et al. 1996). Heelan et al. (1989) examined a series of 31 patients with superior sulcus tumors imaged with both CT and MR and found that MR imaging showed 94% correlation with surgical and clinical findings, whereas the CT scans had an accuracy of only 63% for evaluating tumor invasion through the superior sulcus.

Fig. 14
figure 14figure 14

A 71-year-old male with adenocarcinoma in the right upper lobe. Contrast-enhanced thin-section CT (a) and conventional T1-weighted SE (b) images suggest chest wall invasion, but tumor extent cannot be clearly determined on either image. However, STIR turbo SE (c) image clearly shows chest wall invasion and tumor extent

Full size image

Sakai et al. (1997) used dynamic cine MR imaging during breathing rather than static MR imaging for evaluating chest wall invasion in lung cancer patients. This study evaluated the movement of the tumor along the parietal pleura during the respiratory cycle displayed with a cine loop in a manner similar to dynamic expiratory multi-section CT (Murata et al. 1994). Where the tumor had invaded the chest wall, it was fixed to the chest wall, while without invasion, the tumor was seen to move freely along the parietal pleura. In this study, the sensitivity, specificity, and accuracy of dynamic cine MR imaging for the detection of chest wall invasion were 100%, 70%, and 76%, respectively, and those of conventional CT and MR imaging were 80%, 65%, and 68% (Sakai et al. 1997). Of special significance is that the negative predictive value of dynamic cine MR imaging in this study was 100% without any need for ionizing radiation exposure. Dynamic cine MR imaging, when used in conjunction with static MR imaging, is therefore considered useful for further improvements of the assessment of chest wall invasion in lung cancer patients.

Currently, multiplanar capability, faster scan time and better spatial resolution of thin-section MDCT images may improve the diagnostic capability of CT for evaluation of chest wall invasion in NSCLC patients similar to that of mediastinal invasion. MR imaging is still considered to have superior tissue contrast compared with MDCT. Higashino et al. (2005) also reported that thin-section sagittal MPR imaging with 1 mm section thickness could significantly improve diagnostic accuracy for chest wall invasion in comparison with routine MDCT with 5 mm section thickness and showed slightly better diagnostic capability than thin-section MDCT with 1 mm section thickness. Therefore, further investigations as well as comparative studies of thin-section MDCT imaging and MR imaging used as described here, or of newly developed techniques will be needed to determine the actual significance of MR imaging for assessment of chest wall invasion in routine clinical practice.

4.1.4 MR Assessment of Respiratory Tumor Motion

The general objective of radiotherapy is to achieve tumor control by depositing a lethal dose in the target volume including potential microscopic spread of cancer cells, while sparing surrounding organs and tissue as best as possible. Therefore, precise localization of the target volume is needed. Actually, the recent advances in radiotherapy, including intensity-modulated radiotherapy, adaptive radiotherapy, as well as image-guided radiotherapy, allow for strong improvement of the accuracy of irradiation treatment. However, patient motion and especially respiratory motion has become a major obstacle for achieving high precision radiotherapy. Currently, in classic radiotherapy for lung tumors, such motion is accounted for with a generic target volume expansion, not considering the individual patient breathing characteristics and the mobility of the individual tumor. However, it has been widely recognized that the motion pattern of lung tumors and the breathing cycles vary greatly among patients. This empiric approach includes all surrounding healthy tissues that pass the planned target volume at any time of the breathing cycle to create a safety margin. Within this margin even the additional normal tissues will be irradiated unnecessarily, causing tissue damage or limiting the dose delivered to the target. Therefore it is attractive to define an individual treatment plan by limiting the irradiated volume to certain positions of the tumor on its path during respiration (gated technique) or to follow its respiratory movement (tracking techniques) (Li et al. 2008). Hence, the ultimate objective for radiotherapy of moving targets is to localize precisely the target in space and time in order to achieve a higher dose to be applied to the target while the maximum dose to the normal tissue is reduced, particularly for critical adjacent organs at risk. The better the delineation between target and normal tissue the lower the probability of complications and the higher the chance for tumor control eventually enhanced by the possibility to even increase tumor control by delivering an additional radiation dose solely to the target. For dedicated treatment planning, respiratory-gated four-dimensional (4D) MRI could be used to exactly define tumor size and its three-dimensional displacement during respiration in a single examination. The fourth dimension beyond the 3D space is time, in which patient motion and the change of the position of the tumor will be recorded. Ideally, 4D MRI would not only encode tumor and organ motion information but also provide time-resolved 3D data sets with reduced motion artifacts.

Numerous MR-based investigations of lung and tumor motion in the literature have been limited to examining the motion in a single plane or in a small number of orthogonal planes through the tumor. Two non-coplanar image views provided critical motion characterization, while the most significant displacement was in the cranial-caudal direction (Shimizu et al. 2000). For this purpose, the MR sequences derived from cardiac imaging have been adapted for respiratory motion analysis. These sequences were compared, demonstrating that fast imaging with a free precession steady-state gradient-echo provided significantly higher SNR than any fast low-angle shot technique, while the latter had an advantage in higher temporal resolution.

The correlation between external fiducial markers (coils) and the internal organ motion was also studied using single-slice 4D MRI (Plathow et al. 2005). The correlation coefficients in the three orthogonal directions for different breathing types (thoracic or abdominal) were about 0.8, similar in magnitude to 4D CT. This quantitative information indicates that external fiducial markers might be satisfactory for predicting organ motion.

Volumetric 4D MR imaging was not possible before the recent introduction of multichannel parallel detection systems. Parallel acquisition improves the performance of MR imaging by over an order of magnitude compared to single-channel MR systems. The signal-to-noise ratio (SNR) is usually degraded when using multielement coils for multichannel imaging. Consequently, some of the gain in acceleration is sacrificed in order to maintain image quality. Compared with single-slice imaging, which requires multiple slice directions to view the critical motions of the moving organ, the volumetric 4D technique catches the entire volume in a single acquisition. However, single-slice 4D MR imaging has a higher speed and can be used to study fast heart beating and forced breathing maneuvers. Further improvement of 4D MR imaging may employ the view-sharing technique using a variable sampling rate in k-space and shares elements between image sets, reducing the acquisition time by an appropriate approximation. This technique, combined with parallel imaging, allows for volumetric 4D MR imaging in respiratory motion studies. Then the acquisition time could be below 0.7 s for a 3D torso image using a 1.5-T MRI scanner.

A study directly compared 4D CT, 4D MR imaging, and cone beam CT (CBCT) and demonstrated that lesion sizes were exactly reproduced with 4D CT but overestimated on 4D MRI and CBCT with a larger variability due to limited temporal and spatial resolution (Biederer et al. 2009). In addition, all 4D modalities underestimate lesion displacement (Biederer et al. 2009). In addition, another study demonstrated that this technique was considered as a promising tool to analyze complex breathing patterns in patients with lung tumors and to use in planning of radiotherapy to account for individual tumor motion (Dinkel et al. 2009). Therefore, in the last decades, many investigators have assessed the influence of respiratory motion to radiation therapy as well as PET/MRI by 4D MR imaging as well as different motion correction methods (Dikaios et al. 2012, Tryggestad et al. 2013, Sawant et al. 2014, Dutta et al. 2015, Fayad et al. 2015, Stanescu and Jaffray 2016). Therefore, this part would be better to be more and more important for future lung MR imaging, especially oncologic fields.

4.2 MR Assessment of N Classification

The descriptor N classification refers to the presence or absence of regional lymph node metastases (Sobin and Wittekind 2002). The definitions are given in Table 4. In the absence of distant metastasis, locoregional lymph node spread will determine therapy and prognosis. For patients without positive lymph nodes (N0 disease) or with only intrapulmonary or hilar lymph nodes (N1 disease), direct resection remains standard therapy. In case of positive ipsilateral mediastinal lymph nodes (N2 disease), chemotherapy combined preoperatively with surgery or with concurrent or sequential radical radiotherapy is a legitimate choice (Martini et al. 1997; Vansteenkiste et al. 1998). If patients have contralateral mediastinal lymph node metastases (N3 disease), however, they are generally rejected for surgery but will receive nonsurgical combination treatment.

CT has been the standard noninvasive modality for staging of lung cancer. Enlarged lymph nodes (i.e., with a short axis of more than 10 mm or a long axis of more than 15 mm) are considered to be metastatic. Although an increase in the size of mediastinal lymph nodes correlates with malignant involvement in patients with lung cancer, the sensitivity and specificity of this finding are not very high because lymph nodes can be enlarged due to infection or inflammation. In addition, small nodes can sometimes contain metastatic deposits. The RDOG reported that the sensitivity and specificity of CT for N classification were only 52% and 69%, respectively (Webb et al. 1991), while the corresponding values from the Leuven Lung Cancer Group (LLCG) were 69% and 71% (Dillemans et al. 1994). Due to the substantial limitation of CT for depicting mediastinal lymph node metastases, additional mediastinoscopy with biopsy is necessary for adequate assessment of hilar and mediastinal nodes (Glazer et al. 1984, 1985; Musset et al. 1986; Poon et al. 1987; Laurent et al. 1988; Webb et al. 1991, 1993; McLoud et al. 1992).

Since the 1990s, FDG-PET has been used for differentiation between metastatic and nonmetastatic lymph nodes based on the biochemical mechanism of increased glucose metabolism or tumor cell duplication (Wahl et al. 1994; Patz et al. 1995; Boiselle et al. 1998; Higashi et al. 1998; Gupta et al. 2000). However, elevated glucose metabolism may occur secondary to tumor, infection or inflammation (Dewan et al. 1993; Patz et al. 1993), and spatial resolution in PET is inferior to that of CT and MR, so that the diagnostic capability of the FDG-PET imaging is limited (Gupta et al. 2000). A large number of prospective studies have compared the diagnostic capability of N stage assessment using CT and FDG-PET. A meta-analysis demonstrated that FDG-PET was significantly more accurate than CT for identifying lymph node involvement (Gould et al. 2003). In addition, the respective median sensitivity and specificity of CT were 61% (interquartile range, 50–71%) and 79% (interquartile range, 66–89%), but those of FDG-PET were 85% (interquartile range, 67–91%) and 90% (interquartile range, 82–96%) (Gould et al. 2003). Moreover, it has been suggested that FDG-PET is more sensitive but less specific when CT showed enlarged lymph nodes [median sensitivity, 100% (interquartile range, 90–100%); median specificity, 78% (interquartile range, 68–100%)] than when CT showed no lymph node enlargement [median sensitivity, 82% (interquartile range, 65–100%); median specificity, 93% (interquartile range, 92–100%); (P = 0.002)] (Gould et al. 2003).

Since the introduction of MR imaging for assessment of lung cancer, the criteria for tumor involvement within lymph nodes depend solely on lymph node size and were very similar to CT criteria. In some cases, however, histological examination has shown that a normal-sized regional lymph node may have metastases and that nodal enlargement can be due to reactive hyperplasia or other nonmalignant conditions. The detectability of calcifications, which are indicative for a benign lesion, is also limited for MRI when compared with CT. The direct multiplanar capability of MR imaging, however, is an advantage for the detection of lymph nodes in areas that are suboptimally imaged in the axial plane, such as in the aortopulmonary (AP) window and subcarinal regions (Webb et al. 1991; Boiselle et al. 1998).

Recently, cardiac- and/or respiratory-triggered conventional or black-blood STIR turbo SE imaging has been recommended for detection of metastatic tumors and metastatic lymph nodes (Fujimoto et al. 1995; Eustace et al. 1998; Takenaka et al. 2002; Ohno et al. 2004d, 2007c; Kawai et al. 2006). These novel sequences may make to quantitatively assess the signal intensity of lymph nodes by means of comparison with a 0.9% normal saline phantom (Takenaka et al. 2002; Ohno et al. 2004d, 2007c). The STIR turbo spin echo (SE) is a simple sequence, which can be easily included in clinical protocols to yield net of T1- and T2-relaxation times. On STIR turbo SE images, metastatic lymph nodes exhibit high signal intensity and nonmetastatic lymph nodes low signal intensity (Figs. 15 and 16). According to previously published results (Fujimoto et al. 1995; Takenaka et al. 2002; Ohno et al. 2004d, 2007c, 2011b; Yi 2008; Morikawa et al. 2009), sensitivity of quantitatively and qualitatively assessed STIR turbo SE imaging ranged, on a per-patient basis, from 83.7 to 100.0%, specificity from 75.0 to 93.1%, and accuracy from 86.0 to 92.2%, and these values are equal to or higher than those for CE-CT, FDG-PET, or PET/CT. Yet another study showed that the quantitative and qualitative sensitivity, specificity, and accuracy of STIR turbo SE imaging were not significantly different from those of FDG-PET/CT. However, the combination of FDG-PET/CT with STIR turbo SE imaging was found to be significantly more effective for detecting nodal involvement on a per-patient basis (96.9% specificity, 90.3% accuracy) than FDG PET/CT alone (65.6% specificity, 81.7% accuracy) (Morikawa et al. 2009).

Fig. 15
figure 15figure 15

A 77-year-old male with adenocarcinoma (N0). Integrated FDG-PET/CT (a) demonstrates high uptake of FDG in primary tumor, but no uptake in mediastinal and hilar lymph nodes, suggesting N0. Black-blood STIR turbo SE imaging on axial (b) and coronal (c) planes show the primary tumor as high intensity, but mediastinal (arrow) and hilar (arrow head) lymph nodes as low signal intensity, suggesting N0

Full size image
Fig. 16
figure 16figure 16

A 74-year-old male with adenocarcinoma (N1). Integrated FDG-PET/CT (a) demonstrates no uptake of FDG in the mediastinal and faint uptake in the hilar lymph nodes, suggesting N0, which was false-negative. Black-blood STIR turbo SE imaging on axial (b) and coronal (c) planes shows the hilar lymph node (arrow) as high signal intensity, suggesting N1 which was true-positive

Full size image

Since 2008, DWI was introduced as another promising MR technique for this purpose (Hasegawa et al. 2008; Nomori et al. 2008; Ohno et al. 2011b; Pauls et al. 2012). Sensitivity, specificity, and accuracy of DWI reportedly range, on a per-patient basis, from 77.4% to 80.0%, 84.4% to 97.0%, and 89.0% to 95.0%, respectively, and these results appear to be equal to or better than those for FDG-PET or PET/CT (Hasegawa et al. 2008; Nomori et al. 2008; Ohno et al. 2011b; Pauls et al. 2012). Ohno et al. prospectively and directly compared these modalities to determine the clinical relevance of MR-based N-factor assessment as compared with that of FDG-PET/CT. In this study, sensitivity and/or accuracy of STIR turbo SE imaging (quantitative sensitivity: 82.8%, qualitative sensitivity: 77.4%, quantitative accuracy: 86.8%) proved to be significantly higher than those of DWI (74.2%, 71.0% and 84.4%, respectively) and FDG-PET/CT (quantitative sensitivity: 74.2%) (Ohno et al. 2011b). This means that quantitative and qualitative assessments of the N stage of NSCLC patients obtained with STIR turbo SE MR imaging are more sensitive and/or more accurate than those obtained with DWI and FDG PET/CT (Ohno et al. 2011b). According to these results and considering the limitations of DWI as well as FDG-PET/CT for detection of small metastatic foci or lymph nodes, STIR turbo SE imaging may be the better MR technique to use for this purpose before surgical treatment or lymph node sampling, during thoracotomy or mediastinoscopy for accurate pathologic TNM staging after surgical treatment, or before chemotherapy, radiation therapy, or both (Ohno et al. 2011b). However, further technical improvements in DWI are needed to overcome its current limitations and enable it to function in a complementary role or as a substitution for STIR turbo SE imaging in routine clinical practice. Thus, STIR turbo SE imaging should be considered as capable of enhancing the diagnostic capability of N classification not only due to its multiplanar capability but also its sensitive and accurate assessment of relaxation time differences between metastatic and nonmetastatic lymph nodes and play as complementary and/or substitution of PET or PET/CT as well as DWI in routine clinical practice.

Recently, PET/MRI at 3 T system was introduced for TNM staging in NSCLC patients. Although T staging had no significance with PET/CT, the diagnostic accuracy for N stage of MR imaging including STIR fast advanced spin echo (FASE) imaging, PET/MRI, and PET/CT was 98.6% for MR imaging; 98.6% and 92.1% for PET/MRI with and without signal intensity (SI) assessment based on STIR FASE imaging, respectively; and 92.1% for PET/CT (Ohno et al. 2015). In addition, the accuracy of STIR FASE imaging and PET/MRI with SI assessment was significantly higher than that of PET/MRI without SI assessment and of PET/CT (Ohno et al. 2015). Moreover, sensitivity of STIR FASE imaging (100%) and PET/MRI with SI assessment (100%) were significantly higher than that of whole-body PET/MRI without SI assessment (93.8%) and of PET/CT (93.8%) (Ohno et al. 2015) (Fig. 17). Therefore, PET/MRI may be more useful than PET/CT, when each investigator evaluates SI changes with FDG uptake in not only NSCLC, but also other oncologic patients in routine clinical practice.

Fig. 17
figure 17figure 17

A-66-year-old male with right hilar lymph node metastasis due to adenocarcinoma. CE-thin-section MPR image (a) shows right hilar lymph node (arrow), whose short axis diameter was 6 mm, and diagnosed as N0 case. This patient was false-positive case on CT. PET/CT (b) and PET/MRI fused between PET and contrast-enhanced (CE) Quick 3D data (c) demonstrate small lymph node without high FDG uptake in right hilum (arrow) and also assessed as N0 case. This patient was false-negative case on PET/CT and PET/MRI. Whole-body STIR FASE image (d) shows high signal intensity at right hilar lymph node metastasis (arrow), although whole-body Quick 3D imaging (e) demonstrates small right hilar lymph node (arrow) in the right hilum. Moreover, whole-body MR image fused between diffusion-weighted image (DWI) and CE Quick 3D data (f) demonstrates small right hilar lymph node with low signal intensity color coded as heat (arrow). Therefore, whole-body MR imaging can correctly diagnose this case as N1 case, when applied STIR FASE image.

Full size image

4.3 MR Assessment of M Classification

The descriptor M relates to the presence of distant metastasis (M1) or its absence (M0) (Sobin and Wittekind 2002). The definitions are given in Table 5. Lung cancer can metastasize widely and involve many organs, including the brain, bone, liver, and adrenal glands. The presence of metastasis beyond the intrathoracic lymph nodes is considered an indication of metastatic disease (M1) and implies surgical non-resectability. Patients with distant metastases carry a very poor prognosis and are generally treated with chemotherapy, radiotherapy, or both or with optimal supportive care. In most cases, extrathoracic imaging is indicated for patients with lung cancer and symptoms localized to a specific organ. At present, however, there is no consensus regarding the efficacy of extrathoracic imaging for presumably resectable lung cancer without signs or symptoms localized to a specific organ (Wong et al. 1999).

The observation of metastases in patients with NSCLC has major implications for management and prognosis. Extrathoracic metastases are present in approximately 40% of patients with newly diagnosed lung cancer at presentation, most commonly in the adrenal glands, bones, liver, or brain (Pantel et al. 1996; Quint et al. 1996). After radical treatment for apparently localized disease, 20% of the patients developed an early distant relapse, probably due to systemic micrometastases that were present but not detected or visualized at the point of initial staging (Pantel et al. 1996). Silvestri et al. (1995) updated a meta-analysis for the systemic evaluation of extrathoracic metastases in potentially resectable NSCLC patients. This study calculated that the negative predictive value was equal to or more than 90% for the clinical evaluation of patients asymptomatic for brain, abdominal, or bone metastases (Silvestri et al. 1995). These findings are consistent with the findings of a retrospective analysis of 755 patients with clinical stage T1-2 N0 disease, which found only five sites with silent metastasis after extensive imaging for extrathoracic disease (Tanaka et al. 1999). The current recommendation from the American College of Chest Physicians (ACCP) therefore suggests that further diagnostic testing is necessary to confirm the presence of disease only in patients with abnormal findings on clinical evaluation, although the positive predictive values among the studies included in their meta-analysis were highly variable (Toloza et al. 2003; Silvestri et al. 2007). For purposes of TNM classification, however, it would be necessary to perform in-depth surveillance of potential sites of extrathoracic metastases for all lung cancer patients. In addition, accurate diagnosis of extrathoracic metastases may be helpful for clinicians to provide the most appropriate treatment and/or management for lung cancer patients.

4.3.1 Adrenal Gland Metastasis

Enlarged adrenal glands can be visualized on CT at initial presentation in nearly 10% of NSCLC patients, and approximately two-third of these adrenal lesions are benign or asymptomatic (Oliver et al. 1984; Ettinghausen and Burt 1991). Therefore, without pathologic proof of metastatic disease, the presence of an isolated adrenal mass in a patient with otherwise operable NSCLC should not preclude radical treatment. If the CT scan is performed without intravenous contrast media and an adrenal lesion is identified, measurement of the CT scan attenuation value can be helpful for distinguishing metastasis from adenoma (Boland et al. 1998; Szolar and Kammerhuber 1998). PET can also be a useful adjunct in this setting because the sensitivity and specificity of PET have been reported as ranging from 80 to 100% in the past literatures. This high sensitivity and specificity may result in a reduction of the number of unnecessary biopsies, which are not without risk and not always diagnostic (Erasmus et al. 1997; Marom et al. 1999). However, careful interpretation of PET is required for small lesions less than 10 mm in diameter, since experience with these is still limited (Schrevens et al. 2004). In addition, false-positive findings on PET have also been reported, and the incidence of false-positive findings is increasing. Currently, MR imaging is also considered helpful for distinguishing metastasis from adenoma when an adrenal lesion is detected by CT. Visual assessment of adrenal lesions using chemical shift MR imaging may characterize a lesion as an adenoma on the basis of reduced signal intensity of the lesion on opposed-phase images as compared with that on in-phase images (Korobkin et al. 1995; Schwartz et al. 1998; Hussain and Korobkin 2004). Korobkin et al. (1995) applied this technique to 51 adrenal lesions and reported a sensitivity of 100% and specificity of 81% for the characterization of adenomas.

4.3.2 Bone Metastasis

Bone involvement is usually assessed by 99mTechnetium methylene diphosphate (99mTc-MDP) or hydroxymethylene diphosphate (99mTc-HMDP) bone scintigraphy. Although sensitivity of bone scintigraphy has been reported as high as 90%, its specificity was only about 60% due to false-positive findings caused by the nonselective uptake of the radionuclide tracer in any area of increased bone turnover (Schrevens et al. 2004). Consequently, additional imaging by X-ray, bone CT, and/or MR imaging is often required. PET is reported to have similar sensitivity, but higher specificity and accuracy (equal to or more than 90%, equal to or more than 98% and equal to or more than 96%, respectively) (Bury et al. 1998; Marom et al. 1999). PET is therefore considered superior to bone scintigraphy for the detection of bone metastases. Currently, MR imaging with the use of various sequences such as T1-weighted SE or turbo SE imaging, T2-weighted turbo SE imaging, STIR turbo SE imaging, contrast-enhanced T1-weighted SE or turbo SE imaging, or diffusion-weighted MR imaging is deemed useful for assessment of muscle-skeletal tumors and metastasis from various malignancies (Weinberger et al. 1995; Vanel et al. 1998; Mentzel et al. 2004; Park et al. 2004; Tokuda et al. 2004; Goo et al. 2005). However, only one study has directly compared the diagnostic capability of MR imaging and bone scintigraphy and found that sensitivity, specificity and accuracy of MR imaging were 80%, 96% and 93%, respectively, being superior to bone scintigraphy (40%, 92%, and 83%), although the difference was not significant (Earnest et al. 1999).

4.3.3 Brain Metastasis

Some investigators have reported that brain MR imaging is useful for evaluation of asymptomatic brain metastases in patients with operable lung cancer (Hillers et al. 1994; Earnest et al. 1999; Yokoi et al. 1999). FDG-PET is not suitable for the detection of brain metastases since the sensitivity of PET is low due to the high glucose uptake of normal surrounding brain tissue. CT and/or MR imaging remain the method of choice for screening brain metastases. Yokoi et al. (1999) compared the efficacy of MR imaging and CT scans of brain in 332 patients with potentially operable asymptomatic non-small cell lung cancer. Within 12 months of diagnosis, brain metastases were detected in 7% of the patients in this series. Preoperatively, brain metastases were detected in 3.4% of the patients by MR imaging and in 0.6% of the patients by CT scans. Other investigators have reported on the utility of contrast-enhanced brain MR imaging and found a high prevalence of asymptomatic brain metastasis in 28% of patients identified with contrast-enhanced brain MR imaging (Earnest et al. 1999). These findings suggest that preoperative brain MR imaging may be effective for patients with lung cancer.

4.3.4 Whole-Body MR Imaging for Assessment of M Classification in Lung Cancer Patients

Findings of a recent randomized trial suggest that the addition of whole-body FDG-PET scanning to a conventional workup can identify more patients with extrathoracic metastases among those with suspected NSCLC (van Tinteren et al. 2002). However, recent advances in MR techniques such as fast imaging and moving table techniques make it possible to perform whole-body MR imaging. Its usefulness has been investigated in the staging of breast cancer and the search for primary lesions in patients with metastatic carcinoma from an unknown primary lesion (Eustace et al. 1998; Walker et al. 2000; Antoch et al. 2003; Lauenstein et al. 2004; Takahara et al. 2004; Goehde et al. 2005; Schmidt et al. 2006). It was concluded that whole-body MR imaging may constitute a single, cost-effective imaging test for patients with metastatic carcinoma from an unknown primary (Eustace et al. 1998; Walker et al. 2000; Antoch et al. 2003; Lauenstein et al. 2004; Takahara et al. 2004; Goehde et al. 2005; Schmidt et al. 2006). However, the potential of whole-body MR imaging for lung cancer staging has not yet been satisfactorily delineated. Ohno et al. (2007d) performed a direct comparison of the diagnostic capability of whole-body MR imaging and FDG-PET for the M classification. They reported that the interobserver agreement for whole-body MR imaging was substantially, but not significantly better than for whole-body FDG-PET on a per-site basis and a per-patient basis (Ohno et al. 2007d). For assessment of head and neck metastases, sensitivity (84.6%) and accuracy (95.0%) of whole-body MR imaging were significantly higher than those of FDG-PET (15.4% and 89.1%, respectively) on a per-site basis (Ohno et al. 2007d). In addition, the specificity (96.1%) and accuracy (94.8%) of whole-body MR imaging for bone metastasis were significantly higher than those of FDG-PET (88.3% and 88.2%, respectively) on a per-site basis (Ohno et al. 2007d). However, when brain metastases were excluded from head and neck metastases, sensitivity, specificity, and accuracy of whole-body MR imaging were not significantly different from those of FDG-PET, nor were they for diagnosis of thoracic, abdominal, and pelvic metastases (Ohno et al. 2007d). In addition, when evaluation on a per-patient basis of M classification included brain metastases as head and neck metastases, accuracy (80.0%) of whole-body MR imaging was significantly better than that of FDG-PET (73.3%), while exclusion of brain metastases from head and neck metastases resulted in no significant differences in sensitivity, specificity, and accuracy between whole-body MR imaging and FDG-PET (Ohno et al. 2007d). Whole-body MR imaging is therefore an accurate diagnostic technique and should be considered at least as effective as FDG-PET for M classification of lung cancer patients (Figs. 18 and 19).

Fig. 18
figure 18figure 18

A 68-year-old female with adenocarcinoma and lumbar vertebra (L1) metastasis. Integrated FDG-PET/CT (a) shows high uptake of FDG at the vertebral arch (arrow), suggesting bone metastasis, which was true-positive. On sagittal whole-body MR images (b) (L to R: non-contrast-enhanced T1-weighted GRE image, STIR turbo SE image, and contrast-enhanced T1-weighted GRE image), bone metastasis (arrow) is obvious, which was true-positive

Full size image
Fig. 19
figure 19figure 19

A 67-year-old male with adenocarcinoma and brain metastasis. Integrated FDG-PET/CT (a) does not show any abnormal uptake, suggesting M0, which was false-negative. Contrast-enhanced whole-body T1-weighted GRE (b): L to R, coronal and sagittal planes clearly shows brain metastasis, suggesting M1, which was true-positive

Full size image

Furthermore, whole-body DWI has been recommended as a promising new tool for whole-body MR examination of oncologic patients (64–66). Comparisons of the diagnostic performance of whole-body MR imaging for M-factor assessment with that of FDG-PET or PET/CT have shown that the diagnostic capability of whole-body MR imaging with or without DWI (sensitivity, 52.0–80.0%; specificity, 74.3–94.0%; accuracy, 80.0–87.7%) was equal to or significantly higher than that of FDG-PET or PET/CT (sensitivity, 48.0–80.0%; specificity, 74.3–96%; accuracy, 73.3–88.2%) (Ohno et al. 2007d, 2008b; Yi et al. 2008; Takenaka et al. 2009; Sommer et al. 2012). However, one drawback associated with the use of whole-body DWI in this setting needs to be carefully considered. The specificity (87.7%) and accuracy (84.3%) of whole-body DWI alone on a per-patient basis were significantly lower than those of FDG-PET/CT (specificity, 94.5%; accuracy, 90.4%) (Ohno et al. 2008b). On the other hand, the diagnostic accuracy of whole-body MR imaging combined with DWI (87.8%) was not significantly different from that of FDG-PET/CT, although that of whole-body MR imaging without DWI (85.8%) was lower than that of FDG-PET/CT (Ohno et al. 2008b). Therefore, it would be advisable to use whole-body DWI as part of whole-body MR examination in order to improve the diagnostic accuracy of M-factor assessment of NSCLC patients (Ohno et al. 2008b).

More recently, whole-body PET/MR imaging obtained with a hybrid PET/MR system or presented as PET fused with MR imaging has become clinically feasible, and a few investigators have conducted preliminary tests of their utility for lung cancer patients and reportedly found no significant differences in diagnostic capability between PET/MR imaging and PET/CT for M-factor assessment as well as for T- and N-factor assessments (Kohan et al. 2013; Heusch et al. 2014; Schaarschmidt et al. 2015). However, all these studies assessed only FDG uptake and anatomical information from PET/MR data as well as PET/CT, but not signal intensities on all sites detected on a variety of MR images. Correct evaluation of signal intensity changes at all suspected sites on PET/MR imaging may thus result in better diagnostic performance than PET/CT in the near future, especially when taking into consideration MR results reported during the last few decades, which demonstrated the utility of MR imaging for differentiation of metastatic from nonmetastatic sites in lung cancer patients. A study showed the direct comparison of TNM staging capability among whole-body MRI, PET/MR, and PET/CT (Ohno et al. 2015). This study demonstrated that agreements of assessment of every factor (κ = 0.63–0.97) and clinical stage (κ = 0.65–0.90) were substantial or almost perfect. Regarding capability to assess operability, accuracy of whole-body MRI and PET/MR imaging with signal intensity assessment (97.1%) was significantly higher than that of MR/PET without signal intensity assessment and integrated FDG PET/CT (85.0%, p < 0.001). Therefore, when applying PET/MR in this setting, not only anatomical but also various information from MR part of PET/MR have to be evaluated for improving the potential PET/MR for TNM staging in lung cancer patients.