| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Thoracic Imaging |
1 From the Department of Diagnostic Radiology (F.E., G.M.M., P.H.L., L.A.F., S.J.S.), the Division of Pulmonary and Critical Care Medicine (J.H.R., D.E.M.), and the Section of Biostatistics (C.M.R.), Mayo Clinic, 200 First St SW, Rochester, MN 55905, and the Department of Diagnostic Radiology, Mayo Clinic, Jacksonville, Fla (O.L.B.). Supported in part by grants from the Mayo Foundation for Education and Research and from Bracco Diagnostics. Received November 24, 1997; revision requested February 6, 1998; final revision received August 21; accepted November 5. Address reprint requests to F.E.
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: Twenty-nine patients suspected of having NSCLC localized to the lung or to the lung and regional nodes underwent preoperative MR imaging with contrast material enhancement and radionuclide bone scanning for detection of brain or skeletal metastases. Patients were followed up for 12 months to determine the incidence of clinical metastatic disease.
RESULTS: Eight (28%) patients had occult metastatic disease to the brain or skeleton. Brain metastases were identified on MR images in five of six patients. Bone metastases were identified on MR images in four of five patients and on bone scans in three of five patients. MR imaging was no more accurate than bone scanning for skeletal evaluation. A high dose of MR contrast material allowed detection of more metastases and of small lesions.
CONCLUSION: Contrast-enhanced MR imaging of the brain is indicated for the exclusion of brain metastases in patients with clinically operable known or possible NSCLC and a large (>3-cm) lung mass. Skeletal imaging may be indicated if an isolated brain metastasis is detected.
Index terms: Brain neoplasms, MR, 10.121411, 10.12143 • Brain neoplasms, secondary, 10.3396 • Lung neoplasms, staging, 60.321 • Spine, MR, 30.121411, 30.12143 • Spine, radionuclide studies, 30.12172 • Spine, secondary neoplasms, 30.3396
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The results of early studies (20–22) of MR imaging with contrast material enhancement demonstrated improved detection as compared with the results of nonenhanced MR imaging or CT. The results of comparative studies (23,24) have shown enhanced MR imaging of the brain to be more sensitive than enhanced CT for detection of metastatic disease. The authors of retrospective studies (23,25) of enhanced MR imaging of the brain have suggested even greater sensitivity when a paramagnetic contrast agent is administered at higher doses. Although recent prospective studies (3,26) with CT have been performed, to our knowledge no prospective studies of screening with MR imaging have been reported. In the absence of prospective studies that document the value of "high-dose" gadolinium enhancement, financial constraints mandate that most practices continue to screen for brain metastases by using conventional doses of a gadolinium-based contrast agent.
This prospective pilot study was conducted with the following purposes: (a) to estimate the incidence of occult, preoperative brain and skeletal metastases in patients known to have or suspected of having NSCLC, a lung mass greater than 3.0 cm in diameter (higher stage than T1), and no clinical evidence of distant metastases; (b) to estimate the sensitivity, specificity, and accuracy of MR imaging for help in detecting occult metastases to the brain and of MR imaging of the spine and pelvis and whole-body radionuclide bone scanning for help in detecting occult metastases to the skeleton in these patients; (c) to identify a clinical benefit of a high dose of contrast material for a staging MR examination of the brain; and (d) to confirm positive and negative screening studies at 12-month follow-up.
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Initial evaluation in each patient included acquisition of a complete history and performance of a physical examination, routine laboratory studies, posteroanterior and lateral chest radiography, and CT of the chest and abdomen. Patients with a lung mass larger than 3.0 cm, with no clinical evidence of remote metastases, and with no evidence of mediastinal invasion or abdominal metastases, were eligible for inclusion in the study. Exclusion criteria included pregnancy or lactation; history of lung, breast, or renal cancer in past 5 years; presence of a cardiac pacemaker, cochlear implant, or intracranial aneurysm clip; known history of sensitivity to MR contrast agents; presence of renal failure as demonstrated by a serum creatinine level of more than 2.5 mg/dL (>221 µmol/L); inability to tolerate a curative surgical procedure (pulmonary resection); and inability to provide informed consent for participation.
Results in a comparison group of 110 consecutive patients who subsequently underwent surgical staging and resection of NSCLC without preoperative MR imaging or radionuclide bone scanning were analyzed for selection bias in the study group on the basis of age, sex, tumor histologic findings, and T or N stage.
Preoperative imaging studies for the detection of distant metastases to the brain or skeleton were conducted over a 1–2-day period. A whole-body radionuclide bone scan was obtained with a dual-headed whole-body scanner (Bodyscan; Siemens Medical Systems, Iselin, NJ) equipped with a parallel-hole collimator for each detector head. Approximately 3 hours after intravenous administration of approximately 20 mCi of technetium 99m methylene diphosphonate, whole-body images were obtained with the patient in the supine position. A 20% energy window centered over the 140 keV photopeak was used. Anterior and posterior images were acquired at a speed of 10 cm/min and were formatted with a dual-intensity display on 11 x 14-inch film.
MR imaging was performed with a 1.5-T imager (Signa; GE Medical Systems, Milwaukee, Wis). MR imaging included sagittal T1-weighted spin-echo imaging (repetition time msec/echo time msec, 300/15; two signals acquired, 512 x 512 matrix, 48-cm field of view, 3-mm section thickness, 0.5-mm intersection gap), performed with a phased-array coil, of the cervical, thoracic, and lumbosacral spine and coronal T1-weighted spin-echo imaging (350/16, two signals acquired 256 x 192 matrix, 48-cm field of view, 5-mm interleaved section thickness), performed with a body coil, of the pelvis and proximal portions of the femurs. MR imaging of the brain included sagittal T1-weighted spin-echo imaging (600/16, two signals acquired, 256 x 192 matrix, 24-cm field of view, 5-mm section thickness) performed prior to contrast agent administration, axial T2-weighted spin-echo imaging (2,500/30–90, three-fourths signal acquired, 256 x 192 matrix, 20-cm field of view, 5-mm section thickness, 2.5-mm intersection gap) performed immediately after administration of a conventional dose of 0.1 mmol/kg of gadoteridol (Prohance; Bracco Diagnostics, Princeton, NJ) followed by axial T1-weighted spin-echo imaging (450/16, two signals acquired, 256 x 192 matrix, 20-cm field of view, 5-mm section thickness, 1-mm intersection gap) performed at least 10 minutes after contrast agent administration. A second dose of gadoteridol was then administered, for a total dose of 0.3 mmol/kg of gadoteridol (high dose), which was immediately followed by acquisition of a second series of axial T1-weighted spin-echo images with the same parameters.
Twenty-nine patients successfully underwent radionuclide bone scanning. Twenty-eight patients successfully underwent MR imaging of the axial skeleton and contrast-enhanced MR imaging of the brain with a conventional dose of gadoteridol. One patient did not complete MR imaging with high-dose contrast enhancement of the brain due to transient nausea after administration of the conventional dose. One patient was claustrophobic and was unable to complete any MR imaging examinations.
Two neuroradiologists (G.M.M., P.H.L.) reviewed the MR images, and two nuclear radiologists (L.F.F., O.L.B.) reviewed the bone scans. The reviewing radiologists were informed that the studies had been obtained in patients suspected of having lung cancer. Each radiologist independently interpreted each imaging study and classified abnormalities as (a) probable metastases, (b) indeterminate lesions, or (c) no evidence of metastases. A second interpretation was performed for each study after an interval of at least 3 months, to minimize recall of specific studies. The difference in detectability of enhancing lesions on conventional-dose and high-dose contrast-enhanced MR images was determined by means of consensus. Correlative skeletal radiographs were not available for most patients and were not provided for the blinded interpretation of either the skeletal MR images or the radionuclide bone scans.
The results of 112 MR image interpretations and 116 bone scan interpretations were compared with final pathologic assessment and/or follow-up data and were then evaluated for sensitivity, specificity, and accuracy. For the statistical analyses of diagnostic accuracy (intra- and interobserver agreement), the interpretations were classified as positive (probable metastases) or negative (indeterminate or no evidence of metastases).
The presence of extrathoracic metastases was established by means of clinical and radiologic follow-up for 12 months after initial imaging and/or surgery. Questionnaires were sent to each patient every 3 months after study entry to determine the incidence of clinical metastatic disease to the brain or skeleton. Any follow-up imaging studies of the brain or skeleton in the 29 study patients were obtained and reviewed, if possible. Confirmation of metastatic disease to the brain or skeleton was established by means of biopsy and/or resection results or progressive lesion enlargement demonstrated on successive follow-up CT or MR studies. Negative preoperative imaging studies were considered to be false-negative if a metastatic lesion was detected in the 12 months of follow-up and if the lesion was documented with subsequent imaging studies. Likewise, positive preoperative imaging studies were considered to be false-positive in the absence of histologic proof or if lesion stability or resolution of an imaging abnormality was demonstrated during the 12-month follow-up.
Complete follow-up information through 12 months was available for 26 (90%) of 29 study patients (25 of 27 with NSCLC). Of the three study group patients who did not complete follow-up, one underwent resection of a lung metastasis from colon cancer, another had insufficient pulmonary reserve for pneumonectomy and requested to be removed from the study after 9 months of follow-up, and a third was unable to complete preoperative MR imaging due to claustrophobia; this third patient underwent left upper lobectomy for stage IB (T2N0M0) squamous cell carcinoma and did not complete questionnaires beyond 6 months of follow-up.
Statistical analyses were performed by using the Fisher exact test for comparisons of proportions between groups, the Wilcoxon rank sum test for group comparisons of continuous and ordinal variables, and the 
statistic for the assessment of agreement (27). Interpretations of MR images and bone scans were assessed for sensitivity, specificity, positive and negative predictive values, and accuracy, with the corresponding 95% CIs.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
|
|
|
Eight of 29 patients had evidence of either brain or bone metastases discovered at initial screening examinations and during the 12-month follow-up (Table 4). Occult brain metastases were identified in six of 29 patients (six of 27 with NSCLC) on contrast-enhanced MR images and were verified during 12-month follow-up. Occult skeletal metastases were identified in five of 29 patients (five of 27 with NSCLC) during initial imaging and 12-month follow-up. None of the patients with brain metastases had neurologic symptoms or signs indicative of the presence of brain lesions. Two of five patients with bone metastases had back pain, and one was proved to have a marked elevation of serum alkaline phosphatase level.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Blinded interpretation of the MR studies of the axial skeleton (112 interpretations) yielded 16 true-positive interpretations and 89 true-negative interpretations. False-positive MR studies (n = 3) of bone metastases were obtained in three patients. There also were four false-negative MR image interpretations of bone metastases in a single patient who developed clinical and imaging evidence of metastases 7 months after initial imaging and surgery (Fig 3).
Blinded interpretation of the bone scans (116 interpretations) yielded 10 true-positive interpretations and 91 true-negative interpretations. There were five false-positive and 10 false-negative bone scan interpretations of bone metastases. Again, four of the 10 false-negative bone scan interpretations were in a patient who developed clinical and imaging evidence of metastases 7 months after initial imaging and surgery (Fig 3).
An analysis of intra- and interobserver agreement with the statistic showed almost perfect agreement for interpretation of contrast-enhanced MR studies of the brain (Table 6). There was substantially less agreement for interpretation of skeletal MR and bone scanning studies.
|
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Second, the study was performed prospectively, and, to assess selection bias, the study patients were compared with a group of patients with similar clinical characteristics who did not undergo the screening studies.
Third, contrast-enhanced MR imaging of the brain was used as the screening modality rather than contrast-enhanced CT, which was used in most of the previous screening studies (3–12). MR imaging is more sensitive than contrast-enhanced CT for detection of cerebral metastases (21,24). It is surprising that the superiority of contrast-enhanced brain MR imaging for the demonstration of metastases is not clearly addressed in the current guidelines of the American Thoracic Society and the European Respiratory Society (19).
Fourth, results in the patients were categorized as true-positive or true-negative for brain or skeletal metastases on the basis of their clinical status and subsequent imaging studies obtained during 12 months of follow-up after initial screening. This follow-up was important not only for the accurate assessment of the clinical usefulness of screening examinations for predicting a substantial disease-free interval but also for validation of the accuracy of imaging studies that did not initially demonstrate metastases. The selection of a 12-month follow-up was arbitrary, but a shorter period of 6 months would have altered the data analysis for only one patient (Fig 3).
Occult metastases involving either the brain or skeleton were present in 27% of the 29 patients described in this study and in 30% of the 27 patients with proved NSCLC and a lung mass that was larger than 3.0 cm as measured at chest CT. Preoperative identification of metastases altered therapy and follow-up in all cases. This incidence was consistent with that reported in previous autopsy studies (34) and in long-term follow-up studies with patients with NSCLC (2).
Our results demonstrate a substantial preoperative incidence of cerebral metastases in asymptomatic patients, which was not apparent in many previous studies. The high incidence reported in this study is related both to the use of contrast-enhanced MR imaging for the demonstration of metastases and to the exclusion of patients with a lower risk for extrathoracic metastases (T1N0M0 disease).
By excluding patients with clinical stage IA disease (lung mass diameter 
3 cm), screening was emphasized for patients at greatest risk for asymptomatic metastases and did not include patients who were likely to have either early-stage disease or benign pulmonary lesions. Some patients with T1 pulmonary lesions (3.0 cm) will be included in a screening protocol meant to exclude them. This is due to variation in lesion dimensions as determined at chest CT and pathologic examination of the specimen. Two of the patients in this study had pulmonary lesions slightly smaller than 3.0 cm as measured on the resected specimens. This probably is related to several factors, such as lesion retraction, perilesional infiltrates, and atelectasis, which alter the apparent size of the lesion on chest CT images. Although this study excluded patients with T1 adenocarcinoma, some staging protocols have suggested a screening benefit for these patients, due to an increased incidence of early metastases.
Brain metastases were confidently identified on the initial preoperative screening MR images in six patients and were verified with clinical course or biopsy results. Four of these patients had adenocarcinoma, and two had squamous cell carcinoma. In one patient, a single, small, metastatic lesion, which was apparent only after administration of the high dose of contrast material, was not detected during blinded interpretation of the images on three of four occasions; this lesion was detected at the time of the initial examination and interpretation.
It is interesting to speculate about the variables that made this lesion so difficult to detect at review of the images; factors may include the viewing environment, distraction, fatigue, or the inability to use soft-copy image review with image scrolling and adjustment of window settings. In our daily practice, MR images often are reviewed by scrolling through the images, series by series, at a computer workstation with a large image buffer, with adjustments to the window settings as needed. This "third" dimension is often helpful for differentiating small, focal lesions from cortical vessels. The value of this third dimension with the use of cine spiral CT image display for the differentiation of pulmonary nodules from pulmonary vessels has been previously reported (35). Further studies will be needed to determine the influence of these factors on lesion detection.
The use of MR images enhanced with a high dose of contrast material was of some benefit for the diagnosis of metastatic lesions in three of six patients. A small, solitary lesion was undetectable after routine administration of contrast material, even when approximately 10 minutes were allowed for contrast material diffusion. This lesion became apparent on images obtained with the higher dose. In a second patient, more lesions were detectable after enhancement with a higher dose of contrast material. In a third patient, a lesion was more confidently seen on images enhanced with a higher dose. In only one patient was the use of high-dose contrast enhancement critical to lesion detection and therapy planning. Two of these three patients, who had isolated lesions, underwent stereotactic radiation therapy of presumed brain metastases.
In this study, the usefulness of screening for skeletal metastases was less clear. Both MR imaging and radionuclide bone scanning failed to demonstrate metastatic disease that became apparent both clinically and on follow-up bone scans obtained 7 months after lobectomy for resection of an adenocarcinoma (T1N0M0 disease). In another patient with a single spine metastasis seen on MR images, multiple interpretations of a radionuclide bone scan failed to detect this lesion.
Unlike on contrast-enhanced MR images of the brain, there were three false-positive interpretations of probable metastases on skeletal MR images. Two false-positive interpretations involved areas of heterogeneous-appearing marrow in the pelvis and proximal portions of both femurs. One false-positive MR study and two false-positive bone scans were obtained in a patient with a diffuse marrow abnormality that, at marrow biopsy and 12-month follow-up, proved to represent marrow hyperplasia of unknown cause. In another patient, in whom negative skeletal MR studies were obtained and who had a small pelvic bone island, two bone scans were interpreted as false-positive for metastases. This lesion was unchanged at follow-up bone scanning 9 months after surgery.
The presence of false-positive interpretations is particularly troublesome for a screening examination: Such interpretations can unnecessarily cause a delay in therapy and introduce additional testing to refute the false impression of metastases. Both false-positive and false-negative MR studies of the spine and pelvis undoubtedly are related to the heterogeneity of marrow signal intensity, which can suggest the presence of metastases or mask the presence of small lesions. The difficulty in interpreting bone scans for detection of metastases has been noted in the past (13–18).
There was a moderate statistical association between the presence of brain and bone metastases (Table 4). Acquisition of skeletal images in the six patients with brain metastases would have facilitated detection of bone metastases in three of the five patients with such bone lesions. Acquisition of brain images in the four patients with skeletal metastases would have facilitated detection of brain metastases in three of the six patients with such brain lesions.
This clearly was a pilot study with 29 patients who had advanced NSCLC. The study was intended to help determine a valid approximate preoperative determination of which patients with detectable occult metastases would represent cases previously labeled as "early" postoperative recurrence. Although the estimates of occult metastases are somewhat imprecise due to sample size, several important points are suggested by these data.
1. A substantial number of preoperative patients with NSCLC and a lung lesion larger than 3.0 cm harbor detectable occult brain metastases. Patients who have been determined to have metastases may need additional therapy at the time of initial treatment, modified follow-up, or reconsideration of surgical therapy for the primary tumor.
2. Patients without brain metastases after preoperative contrast-enhanced brain MR imaging did not develop clinical evidence of brain metastases after 12 months of follow-up.
3. Contrast-enhanced brain MR studies were interpreted reliably, and there seems to be a very low risk of false-positive interpretation of these studies.
In conclusion, preoperative contrast-enhanced MR imaging of the brain should be performed in patients with large pulmonary lesions (>3-cm diameter or higher than T1 stage) that are suggestive of NSCLC and can demonstrate asymptomatic cerebral metastases in a substantial number of patients who harbor such lesions. The determination of which patients have occult brain metastases is essential for the specification of appropriate treatment and follow-up testing. Preoperative screening of patients for skeletal metastases is less accurate and may yield both false-positive and false-negative studies. In retrospect, two of the five study patients with skeletal metastases had bone pain ascribed to arthritis, and one of these patients also had a clinically important elevation of serum alkaline phosphatase level. Because of the association of brain and skeletal metastases, preoperative MR imaging of the spine and pelvis should be performed in patients who demonstrate evidence of a single brain metastasis, if surgical resection of the brain lesion is considered (36).
A prospective, multiinstitutional, randomized study is necessary to further refine the benefits and costs of preoperative MR imaging to screen for brain metastases in patients with higher than T1 stage NSCLC and to define the benefits of administration of a high dose of contrast material.
|
Footnotes |
|---|
Author contributions: Guarantor of integrity of entire study, F.E.; study concepts and design, F.E., J.H.R., S.J.S.; definition of intellectual content, F.E., J.H.R.; literature research, F.E.; clinical studies, F.E., J.H.R., G.M.M., P.H.L., L.A.F., O.L.B., D.E.M.; data acquisition, F.E.; data analysis, F.E., C.M.R.; statistical analysis, F.E., C.M.R.; manuscript preparation, F.E.; manuscript editing and review, F.E., J.H.R., G.M.M., P.H.L., L.A.F., O.L.B., C.M.R., S.J.S., D.E.M.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  Y. Ohno, H. Koyama, Y. Onishi, D. Takenaka, M. Nogami, T. Yoshikawa, S. Matsumoto, Y. Kotani, and K. Sugimura Non-Small Cell Lung Cancer: Whole-Body MR Examination for M-Stage Assessment--Utility for Whole-Body Diffusion-weighted Imaging Compared with Integrated FDG PET/CT Radiology, August 1, 2008; 248(2): 643 - 654. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. A. Silvestri, M. K. Gould, M. L. Margolis, L. T. Tanoue, D. McCrory, E. Toloza, and F. Detterbeck Noninvasive Staging of Non-small Cell Lung Cancer: ACCP Evidenced-Based Clinical Practice Guidelines (2nd Edition) Chest, September 1, 2007; 132(3_suppl): 178S - 201S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Lardinois, W. Weder, M. Roudas, G. K. von Schulthess, M. Tutic, H. Moch, R. A. Stahel, and H. C. Steinert Etiology of Solitary Extrapulmonary Positron Emission Tomography and Computed Tomography Findings in Patients With Lung Cancer J. Clin. Oncol., October 1, 2005; 23(28): 6846 - 6853. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Erturan, M. Yaman, G. Aydin, I. Uzel, B. Musellim, and K. Kaynak The Role of Whole-Body Bone Scanning and Clinical Factors in Detecting Bone Metastases in Patients With Non-small Cell Lung Cancer Chest, February 1, 2005; 127(2): 449 - 454. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. M. Finke, M.-C. Aubry, H. D. Tazelaar, G. L. Aughenbaugh, C. M. Lohse, V. S. Pankratz, and C. Deschamps Autopsy Results After Surgery for Non-Small Cell Lung Cancer Mayo Clin. Proc., November 1, 2004; 79(11): 1409 - 1414. [Abstract] [PDF]
|
|
|
|
 |
|
 R. E Benamore, J. J Entwisle, and M. D Peake Bone scanning in lung cancer: Evidence is not sufficient to justify routine bone scanning BMJ, July 24, 2004; 329(7459): 230 - 231. [Full Text]
|
|
|
|
|
|
D. G. Pfister, D. H. Johnson, C. G. Azzoli, W. Sause, T. J. Smith, S. Baker Jr, J. Olak, D. Stover, J. R. Strawn, A. T. Turrisi, et al. American Society of Clinical Oncology Treatment of Unresectable Non-Small-Cell Lung Cancer Guideline: Update 2003 J. Clin. Oncol., January 15, 2004; 22(2): 330 - 353. [Full Text] [PDF]
|
|
|
|
 |
|
 S. G. Spiro and J. C. Porter Lung Cancer--Where Are We Today?: Current Advances in Staging and Nonsurgical Treatment Am. J. Respir. Crit. Care Med., November 1, 2002; 166(9): 1166 - 1196. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 British Thoracic Society, Society of Cardiothoracic Surgeons of Great Britai, and Ireland Working Party Guidelines on the selection of patients with lung cancer for surgery Thorax, February 1, 2001; 56(2): 89 - 108. [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
