Lung cancer staging: the role of computed tomography and magnetic resonance imaging.

Because complete resection remains the only reliable method of cure of lung cancer, one important aim of preoperative staging is to select patients with localised disease who may benefit from surgery, while avoiding unnecessary thoracotomies in patients with unresectable neoplasm. Computed tomography (CT) of the chest is a valuable method for staging local and regional spread of lung neoplasms, although limitations in its accuracy are well-known. While gross invasion of the mediastinum and major structures as well as the presence of metastatic disease can be easily demonstrated with CT, differentiation between tumour contiguity and subtle invasion of mediastinum or chest wall often remains a problem. Although magnetic resonance imaging (MRI) may have the same limitations as CT, in specific situations it may be superior in diagnosing minimal chest wall or mediastinal invasion. Moreover, MRI is useful in the assessment of patients with superior sulcus tumours as well as in patients with contraindication to intravenous administration of ionic contrast material. Since nodal size is the only useful criterion for evaluating lymph node metastases, CT and MRI show similar, poor accuracies in lymph node staging resulting from both low sensitivity (normal-sized nodes may contain microscopic metastases) and low specificity (enlarged lymph nodes may be reactive). For this reason, if enlarged lymph nodes are detected, further evaluation is recommended before excluding the patient from a potentially curative resection. Advantages and limitations of CT and MRI in the preoperative staging of non-small-cell carcinoma are reviewed in this article. The imaging of small-cell carcinoma is not included because most patients with this cell type do not benefit from surgical resection. Similarly we do not discuss imaging of distant metastases.

Spiral CT of the mediastinum: optimization of contrast medium use.

Spiral computed tomography (CT) allows very satisfactory temporal resolution for the detection and analysis of changes in contrast medium density, so that volumes of contrast material smaller than those recommended for conventional CT can be used. The present double-blind, parallel group study was designed to compare image quality and diagnostic efficacy obtained with two different iodine strengths (200 and 300 mgI/ml) and two different flow rates (2 and 3 ml/s) of a fixed volume (70 ml) of iomeprol in adult consenting patients requiring spiral CT of the mediastinum. Imaging was performed during suspended respiration and, a 15-s delay to scan. Spiral CT was initiated 1 cm above the aortic arch and continued inferiorly for 24 cm in all patients. Two independent readers blindly graded image quality, opacification of the superior vena cava, thoracic aorta and pulmonary arteries, and overall diagnostic quality of the CT examination. CT density measurements were performed over the level of the aortic arch, left and right pulmonary arteries, left atrium, and descending aorta. Opacification of the mediastinal vessels was better in the patients injected at 3 ml/s flow rate and was independent of the iodine strength used, except in the case of the thoracic aorta, which was better opacified by injecting iomeprol 300 mgI/ml at 3 ml/s. The highest diagnostic efficacy of the spiral CT examination was obtained with iomeprol 300 mgI/ml at 3 ml/s infusion rate, which seems to represent the administration scheme of choice.

Neoplastic extension across pulmonary fissures: value of spiral computed tomography and multiplanar reformations.

To assess the value of spiral computed tomography (CT) with multiplanar reformations for detection of neoplastic extension across pulmonary fissures, 51 patients with a lung neoplasm near a fissure underwent spiral CT, followed by multiplanar reformations, and spaced thin-section CT scans through the area of contact between tumor and fissure. The CT studies were evaluated for visibility of fissures and their relationship to the tumor. Imaging findings were compared with surgical results in 31 patients who underwent thoracotomy. Visibility of fissures on multiplanar reformations was either good or acceptable in 47 (92.2%) patients, and poor in four. Surgical findings of neoplastic extension across the major or the minor fissure were present in seven and eight patients, respectively. Thin-section CT scans were 83.3% sensitive in assessing neoplastic involvement of the major fissure, axial CT scans were 57.1% sensitive, and spiral CT multiplanar images were 100% sensitive. In the evaluation of the minor fissure, thin-section CT and axial spiral CT scans were considered inconclusive in six patients whereas multiplanar reformations enabled correct assessment of the fissure/neoplasm relationship in all but one patient. Spiral CT multiplanar images are accurate for detection of transfissural neoplastic extension, and are superior to axial CT scans for evaluation of tumors near the minor fissure.

[Neoplastic involvement of pulmonary fissures. Diagnostic possibility of high resolution x-ray computed tomography]. / Atteinte néoplasique des scissures pulmonaires. Possibilité diagnostique de la tomodensitométrie en haute résolution.

Acknowledging fissure invasion by pulmonary tumors located in contact with or in the vicinity of a fissure is an important piece of information for the therapeutic choice, especially in patients whose functional respiratory impairments are a contraindication of pneumonectomy. Fifteen patients with pulmonary neoplasms adjacent to a fissure were studied with standard computed tomography (10 mm sections), completed by high-resolution CT. The findings of the CT studies were compared with those operative reports. On surgery, the oblique fissure was infiltrated in 12 patients, unharmed in 2 others, while the horizontal fissure was infiltrated in 3 patients. Standard CT allowed diagnosing the involvement of the oblique fissure in 4 patients. Thin sections with high-resolution reconstruction algorithms allowed detecting the involvement of the oblique fissure in 13 cases, including one false-positive result. On thin sections, the fissures appear as well-delineated, dense lines. This allows an accurate study of the relationships between the tumor and the fissure, thus increasing the sensitivity of CT for the detection of tumoral extension across the fissures. The orientation of the horizontal fissure, which is almost parallel to the plane of section, makes the study of its relationships with an adjacent mass difficult, even in high-resolution CT.

Pulmonary embolism: the role of computed tomography and magnetic resonance imaging.

The technological progress in Computed Tomography (CT) (spiral and electron beam) and Magnetic Resonance Imaging (MRI fast sequences) has stimulated their interest in the diagnosis of acute and chronic pulmonary embolism (PE). They are noninvasive procedures able to identify thrombi up to the level of segmental pulmonary branches. This result, albeit not ideal, is significant, in view of the lower clinical relevance of peripheral emboli as compared to more central locations, especially in the absence of peripheral venous thrombosis. Spiral CT allows satisfactory assessment of pulmonary branches with high sensitivity (65-100%), specificity (89-96%), positive predictive value (95%) and negative predictive value (80-100%) in the diagnosis of PE. MRI with spin-echo sequences has also a satisfactory sensitivity (75-90%), specificity (up to 100%), positive predictive value (86%) and negative predictive value (85%). Recently, MR angiography was shown to be able to depict smaller pulmonary branches (6th and 7th generation), even if its efficacy in the identification of emboli has not been demonstrated as yet. CT and MRI are bound to play an increasingly relevant role in the diagnosis of PE.

Contrast enhanced ultrasonography (CEUS) in peripheral lung lesions: A study of 60 cases.

INTRODUCTION: The authors report their experience in 60 patients with infectious and neoplastic peripheral pulmonary lesions studied by conventional radiology, B-Mode ultrasound (US) and computed tomography (CT). In view of the particular pulmonary vascularization (consisting of both pulmonary and bronchial arteries) the patients underwent also contrast enhanced ultrasound (CEUS) using a II-generation contrast agent, SonoVue (sulphur hexafluoride microbubbles surrounded by a phospholipid shell). METHODS AND RESULTS: In this study, the sensitivity of CEUS reached 95% in the characterization of peripheral pulmonary lesions, which is similar to the sensitivity of CT (97%). The method used in this case-study was free of significant side effects. DISCUSSION: This preliminary clinical experience seems to confirm the possibility of using SonoVue enhanced US to make a differential diagnosis between infectious and neoplastic lesions based on a qualitative and quantitative assessment, by evaluating the enhancement pattern (homogeneous or inhomogeneous), arrival time of the contrast agent in the lesion, the possibility to identify the pulmonary arteries and time of contrast agent elimination.

[Radiological examination of the thorax in intensive therapy centers]. / L'esame radiologico del torace nei Centri di terapia intensiva.

Bedside chest radiography plays a considerable role in the whole of the exams carried out in hospitals, especially in intensive care units. Many clinical problems (pulmonary, cardiac, and mediastinal) are related to this examination and can be dealt with only when the radiologist is provided with high-quality radiographs. Therefore, it is extremely important for the radiologist himself to be aware of the various factors implied in the making and reproducing of bedside chest radiographs, which is a fundamental step in the monitoring of the patients undergoing intensive care. These factors can be divided into methodological factors (patient positioning, focus-film distance, and exposure) and technical factors (portable units, survey system). In this paper, a special emphasis is put on a survey system consisting in double differentiated screens and double film, which allows the acquisition of a normo-exposed radiograph of the parenchyma and of one of the mediastinum with a single exposure. Besides these technical and methodological elements, the radiologist must pay attention to the diagnostic-therapeutical means (such as tubes, small probes, and catheters) used on most patients in intensive care units. Their exact positioning must always be checked, and possible incorrect positioning and complications must be detected promptly. All these factors, as well as their scrupulous and constant application, contribute to markedly improve the dialogue between clinician and radiologist, by allowing a rapid evaluation of the cause of the clinical case under examination and its adequate treatment.

[Volumetric computerized tomography (spiral CT). Technical principles and possible applications in the thorax]. / Tomografia computerizzata volumetrica (spiral TC). Principi tecnici e possibili applicazioni nel torace.

Spiral volumetric CT is a new volume scanning technique allowing the one-acquisition imaging of an anatomical volume, by advancing the patient through a continuously rotating X-ray tube. As a result, the X-ray tube moves in spirals around the patient and this is the reason why this new technique is called Spiral CT. At the end of the acquisition, planar images are reconstructed by means of a dedicated image reconstruction algorithm with a linear interpolation between two contiguous spiral segments. Spiral CT offers many advantages, all resulting from the possibility of evaluating anatomical volumes in a short time: the organs which are subject to respiratory motion can be studied in a single breath-hold scan, the quality of secondary reconstructions (2/3 dimensional) is improved, the use of contrast media is optimal and finally motion artifacts are reduced. The authors report their early experience with spiral CT in the chest. Spiral CT is more reliable than conventional CT in the evaluation of pulmonary nodules, which can be easily identified with no risk of missing small lesions, and of the pulmonary apex and of the diaphragm. Moreover, contrast media can be used at a lower dosage than usually recommended with conventional CT.

[Peripheral pulmonary neoplasms: combined imaging (computerized tomography, ultrasonography, magnetic resonance) in the evaluation of chest wall involvement]. / Neoplasie polmonari periferiche: imaging integrato (tomografia computerizzata, ecografia, risonanza magnetica) nella valutazione del coinvolgimento parietale.

Chest wall involvement by a peripheral lung neoplasm, found in 4-8% of the cases, makes a difference in both staging and surgical approach. In the evaluation of the regional extent of the tumor, the radiologist must give the surgeon as much information as possible about relationships between lung neoplasm and chest wall structures, to correctly evaluate stage-T2 cancers and differentiate them from stage-T3 cancers. The authors report the results of a study made to assess the value of Computed Tomography (CT), ultrasonography (US) and Magnetic Resonance Imaging (MRI) to define chest wall invasion by a peripheral bronchogenic carcinoma. To this purpose 25 patients with primary peripheral lung neoplasm were studied with CT, US and MRI and the results were compared with surgical findings by means of TNM classification. Overall sensitivity and specificity were 68% and 74% for CT, 74% and 72% for US and 88% and 86% for MRI, respectively. CT, which still plays the major role in staging lung cancer, seems to get the worst results because of its impossibility to differentiate the cancers adjoining the chest wall from those showing signs of initial invasion. In these cases the use of the other imaging techniques (US, MRI) depends on their availability and on the specific experience of the radiologist.

[Correlation of traditional radiology and echography in circumscribed pleural opacities]. / Correlazione tra radiologia tradizionale ed ecografia nelle opacità pleuriche circoscritte.

The value of ultrasounds in the location of small pleural opacities was investigated. A correlation between conventional radiological and echographic signs was made to define the echographic signs useful in the identification of peripheral pleural and parenchymal lesions.

[Bedside thoracic radiography: a comparison between 3 different types of grid]. / Radiografia del torace a letto: confronto fra tre differenti tipi di griglia.

Bedside chest radiography accounts for an increasingly large portion of all chest X-ray examinations. Nevertheless, image quality is often poor mainly because of scattered radiations which decrease image contrast. Moreover, usually no grid is employed because of difficult beam alignment. This work was aimed at comparing different radiologic grids for bedside chest radiography. Fifty patients submitted to two bedside chest radiographs in 24 hours were studied. All the patients underwent the first exam with a Kodak InSight cassette with a newly-designed (columnar type) grid inside, while the second exam was performed with a conventional 8:1 focused lead-strip grid (Gilardoni) in 25 patients and with a 6:1 focused lead-strip grid (Gilardoni) in the extant 25 patients. Both grids were assembled in a Kodak InSight radiographic cassette. Three independent radiologists evaluated film quality, focusing on the depiction of some anatomical structures--e.g., the tracheobronchial tree, the retrocardiac lung, and devices. Seventy-five examinations were obtained for each grid and for each evaluated structure and graded as "good", "acceptable" and "poor". In the 6:1 vs columnar grid test, the highest rate of "good" and "acceptable"--i.e., diagnostic--findings was observed with the columnar grid in evaluating pulmonary vessels (71/75); the highest rate of "poor" findings was obtained with the same grid in evaluating tracheal bifurcation (43/75). In the 8:1 vs columnar grid test, the highest rate of diagnostic findings was shown by the conventional grid in evaluating retrocardiac lung parenchyma and by the columnar grid for pulmonary vessels (69/75); the highest rate of "poor" findings was obtained with the columnar grid in evaluating tracheal bifurcation (40/75). The statistical analysis of the results (Wilcoxon test) was made to compare the two conventional grids with the new columnar one. Statistically significant differences were observed between the 8:1 grid and the columnar grid to evaluate the bronchial tree. No differences were observed between the 8:1 grid and the columnar grid. Furthermore, to determine the effects of different degrees of grid decentering on image quality, a series of exposures was made using a lung-chest phantom. The grids, both the lead-stripe and the columnar one, were comparable. The higher-ratio grid proved better in evaluating tiny details. The columnar grid exhibited better tolerance to X-ray beam and to grid decentering.

[Staging of pulmonary apex tumors. Computerized tomography versus magnetic resonance]. / Stadiazione dei tumori dell'apice polmonare. Tomografia computerizzata versus risonaza magnetica.

Apical lung cancers account for about 5% of pulmonary lesions and can be divided into two groups: Pancoast and non-Pancoast lesions. Recently, the use of MRI has been suggested in combination with CT to stage this kind of lung cancer. In this paper the authors' experience is reported relative to the current role of MRI and CT in the staging of apical lung cancers. Twelve male patients (mean age: 60.5 years) with apical lung cancers underwent conventional X-ray, CT and MR examinations of the chest. CT and MR images were studied by two independent radiologists with specific experience; surgery was the gold standard in three patients and MR and clinical symptoms in the patients not referred for surgery. In 15/108 cases (13.8%) CT and MR findings were in disagreement but in 93/108 cases (86.2%) they were in agreement. The highest disagreement rate was observed in the study of apical chest wall infiltration (33.3%), while in the study of anonymous vein involvement CT and MRI were always in agreement. The correct assessment of the regional extent of apical lung cancers is mandatory for treatment planning. In this kind of tumors MRI can be considered the method of choice thanks to its high contrast resolution and multiplanar imaging capabilities.

[Magnetic resonance angiography of the pulmonary artery. Preliminary considerations]. / Angiografia a risonanza magnetica dell'arteria polmonare. Considerazioni preliminari.

The possibility was evaluated of imaging the pulmonary artery with MR angiography. Twenty healthy volunteers were studied using 3D FT gradient-echo sequences on the coronal plane, with post-processing by the maximum intensity projection method. TE and TR remaining short, flip angles were selected to increase pulmonary artery signal in contrast with hypointense adjacent tissues and vessels. Flip angle selection allowed the optimal differentiation between pulmonary artery and aorta with 15 degree-25 degree angles (range: 110.7 to 122 for the 15 degree flip angle and 158.7 to 182.1 for the 20 degree flip angle). The sequence was obtained on the coronal plane and the following parameters were employed: TR 0.03 s, TE 10 ms, flip angle 15 degree-20 degree, slice of the total volume 100 mm with 64 partitions, 256 x 256 matrix, 1 zoom factor, 1 acquisition. The patient was positioned with the right hemithorax raised by 30 degrees to visualize the common pulmonary artery and lying on his back, face upward, to visualize the right and left pulmonary arteries. Post-processing employed axial plane rotations from -45 degrees to +45 degrees, with 5 degrees step, and from 0 degrees to 180 degrees, with 15 degrees step. Angio-MR images of the pulmonary artery allowed the visualization of its main components, up to its right and left lobar branches. The main limitation of this technique consisted in its poor spatial resolution.

[Magnetic resonance imaging of the brachial plexus. Anatomy]. / Imaging con Risonanza Magnetic del plesso brachiale. Anatomia.

The brachial plexus is a difficult region to evaluate with radiological techniques. MR imaging has great potentials for the depiction of the various anatomical structures of the brachial plexus--i.e., spinal ganglion, ventral nerve rami root exit of the neural foramina, trunks and cords. Moreover, MR imaging, thanks to its direct multiplanarity, to its excellent soft-tissue contrast, and to its lack of motion artifacts, allows good evaluation of pathologic conditions in the brachial plexus, especially traumas and cancers. On the contrary CT, in spite of its high spatial resolution and good contrast, cannot demonstrate the anatomical structures of the brachial plexus. US detects superficial structures, and conventional radiographs depict only indirect changes in the adjacent lung apex and skeletal structures. From November 1989 to May 1990, 20 normal volunteers (15 males and 5 females; average age: 35 years) were studied with MR imaging. Multisection technique was employed with a dedicated coil and a primary coil. The anatomical structures of the brachial plexus were clearly demonstrated by T1-weighted sequences on the sagittal and the axial planes. T2-weighted pulse sequences on the coronal plane were useful for the anatomical definition of the brachial plexus and for eventual tissue characterization. The correct representation of the anatomical structures of the brachial plexus allowed by MR imaging with our standard technique makes MR imaging the most appropriate exam for the diagnosis of pathologic conditions in the brachial plexus, although its use must be suggested by specific clinical questions.

[Neoplastic involvement of pulmonary fissures. Usefulness of high resolution CT]. / Coinvolgimento neoplastico delle scissure polmonari. Possibilità della TC ad alta risoluzione.

In order to plan an adequate treatment, it is very important to recognize the transfissural spread of lung cancer near a pleural fissure, especially in the patients who cannot bear pneumonectomy because of impaired pulmonary function. Fifteen patients with lung cancer near a fissure were examined by means of conventional CT (10 mm-thick slices) followed by high-resolution CT (HRCT). The results obtained with both techniques were compared with surgical findings. At surgery, major fissure involvement was seen in 12 cases, minor fissure was involved in 3, and 2 patients were normal. In 4 cases fissural involvement was correctly demonstrated by conventional CT scans. Neoplastic involvement of the major fissure was identified in 13 cases by HRCT scans, with only 1 false positive. Pleural fissures appear as thin and sharp lines on HRCT scans. Thus, the relationship of lung cancers to fissures is better identified and the accuracy of the method in defining neoplastic spread is higher. However, the relationship of a neoplastic mass to the minor fissure is difficult to evaluate, because the fissure is roughly parallel to the scanning plane.

[Bronchogenic carcinoma staging: comparison of magnetic resonance/computerized tomography]. / Stadiazione del carcinoma broncogeno: confronto risonanza magnetica/tomografia computerizzata.

The staging of bronchogenic carcinoma is an important factor to select the appropriate treatment. Indeed, the definition of locoregional spread and of hilar and mediastinal node involvement is essential for both correct preoperative assessment and prognostic evaluation of bronchogenic carcinoma. CT and MR imaging are the methods of choice in the evaluation of T and N, even though other techniques--e.g., US and nuclear medicine--can also provide valuable diagnostic information. The authors examined 50 patients with primary bronchogenic carcinoma by means of plain radiographs, CT and MR of the chest. MR and CT findings were compared with surgical results on the basis of TNM classification. In the evaluation of T, sensitivity and specificity were 66% and 88.6%, respectively, for CT and 76% and 92% for MR imaging. No statistically significant differences were found between the two imaging methods (p = 0.19). In the evaluation of N, sensitivity and specificity were 56% and 78%, respectively, for CT and 76% and 88% for MR imaging. A statistically significant difference was found between MR and CT accuracy rates (p = 0.934). CT and MR results were in disagreement especially in the evaluation of hilar lymph nodes. To date, MR imaging cannot be considered a substitute for or a routine adjunct to CT in the staging of bronchogenic carcinoma due to its poor spatial resolution, to the presence of artifacts (especially with high-intensity fields), its cost and limited availability. However, in the evaluation of specific anatomical regions--e.g., the pulmonary apex and the peridiaphragmatic region--MR can provide more diagnostic information than CT thanks to its multiplanarity and better contrast resolution.

[Possibility of false negative results in the diagnosis of lung neoplasms with computerized tomography]. / Possibilità di falsi negativi nella diagnosi delle neoplasie polmonari con Tomografia Computerizzata.

PURPOSE: To describe the radiological appearance of overlooked malignant pulmonary lesions at CT and to analyze the reasons of misdiagnosis. MATERIAL AND METHODS: Ten patients with pulmonary malignant lesions (PML) overlooked at first CT examination were selected among patients with lung cancer who were referred to our institution between 1994 and 1997. CT scans were evaluated by consensus of two chest radiologists with different experience in chest radiology, who were blinded to the final diagnosis. RESULTS: The overlooked pulmonary lesions were 5 endobronchial cancers and 5 central solitary nodules. The mean diameter of the lesions ranged 1 to 2 cm. Furthermore, 7 patients had associated pleural and/or parenchymal abnormalities. The technical quality of CT examinations was considered good in 2 cases, acceptable in 6 cases, poor in 2 cases. PML were correctly identified in 6/10 cases by the first (more experienced) radiologist and in 4/10 by the second radiologist. Four cases were considered suspect for the presence of PML by the first radiologist, 2 by the second. CONCLUSIONS: Endobronchial location of the tumors and their small size were the most frequent causes of misdiagnosis of PML at chest CT in our series. However, a systematic evaluation of CT scans can reduce the percentage of missed lesions.

[Thoracic radiography. Comparison of asymmetric screen-film, technic of mechanical homogenization, and conventional technics]. / Radiografia del torace. Confronto tra sistemi di rilevazione asimmetrici, tecnica di omogeneizzazione meccanica e tecnica tradizionale.

This study was aimed at comparing three different systems, i.e., asymmetric screen-film, mechanical homogenized and conventional techniques, in standing frontal and lateral plain radiographs of the chest. Two hundred consecutive patients with normal posteroanterior (PA) and lateral films were randomly subdivided into 4 groups. Each group was submitted to chest radiographs with a different technique: asymmetric screen-film systems (InSight HC and High Light GCA), mechanical homogenized (Tau-Gil Homogenized) and conventional high-kilovoltage techniques. The exposure values for frontal projections ranged 110 kV (InSight HC) to 141 kV (Tau-Gil), while for lateral projections they ranged 123 kV (conventional technique) to 143 kV (Tau-Gil). Statistically significant differences were observed between the two asymmetric systems as regards exposure values, High Light exhibiting higher mean values in the frontal projection (t-test p < 0.05). Image quality was studied jointly by 3 experienced chest radiologists. The observers were asked to grade, on a 3-point ordinal scale, the conspicuity of mediastinal borders, of pulmonary vessels and of selected areas of lung parenchyma (i.e., retrocardiac, retrosternal and apical regions), as well as overall image quality on the frontal projection. The statistical analysis of paired differences was performed with the Mann-Whitney U-test. The asymmetric and the mechanical homogenized techniques were much better than the conventional technique in depicting tracheobronchial tree, retrocardiac parenchyma, azygos-esophageal recess and thoracic spine (p < 0.05). The mechanical homogenized system provided best overall image quality on frontal films, being superior to both InSight HC and conventional techniques, but not to the High Light GCA system; only the frontal projection obtained with the homogenized technique was compared, no filter being available for the lateral projection. When the two asymmetric systems were compared, the High Light system better depicted vascular structures on frontal films (p < 0.05), while apical areas were better demonstrated with the InSight system, namely with lateral films (p < 0.05).