| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Genitourinary Imaging |
1 From the Joint Department of Medical Imaging, University Health Network and Mount Sinai Hospital, University of Toronto, 610 University Ave, Toronto, ON, Canada M5G 2M9 (M.A.H., S.G., K.J.); and Department of Biostatistics, University Health Network and Princess Margaret Hospital, Toronto, Ontario, Canada (G.L.). Received April 28, 2003; revision requested July 9; final revision received September 23; accepted November 17. Address correspondence to M.A.H.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Adrenal MR images obtained from January 1998 to February 2003 were reviewed. Patients were excluded if they did not undergo unenhanced computed tomography or did not have an adrenal mass with attenuation higher than 10 HU, adequate follow-up, or pathologic diagnosis for use as a reference standard. A diagnosis of adenoma required at least 24 weeks of stability on images. Thirty-eight masses in 36 patients were identified (27 adenomas, nine metastases, one adrenocortical oncocytoma, and one pheochromocytoma). Signal intensity (SI) decrease between in-phase and opposed-phase MR images was measured for the entire mass and normalized to the renal parenchymal SI. In 21 of 36 (58%) patients, dual-echo single–breath-hold MR imaging was used to eliminate misregistration.
RESULTS: The attenuation of 61% (23 of 38) of all masses and 70% (19 of 27) of adenomas was 10–30 HU. With a threshold of 20% SI decrease, the sensitivity of chemical shift MR imaging for hyperattenuating adenoma was 67% (18 of 27 masses). When considering masses with attenuation of 10–30 HU, the sensitivity for adenoma was 89% (17 of 19 masses) and remained reasonable at 75% (six of eight masses) for adenomas with attenuation of 20–30 HU. Only one adenoma with attenuation higher than 30 HU had SI decrease of more than 20%. Specificity for diagnosis of adenoma was 100% (11 of 11).
CONCLUSION: In certain circumstances, chemical shift MR imaging is a reasonable second imaging test for further characterization of a hyperattenuating adrenal mass.
© RSNA, 2004
Index terms: Adrenal gland, CT, 86.12112, 86.12115 • Adrenal gland, MR, 86.121411, 86.121414 • Adrenal gland, neoplasms, 86.317, 86.33 • Computed tomography (CT), phase imaging • Magnetic resonance (MR), chemical shift, 86.121414 • Pheochromocytoma, 86.328
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Washout CT has some disadvantages. In current practice, radiologists are often unable to monitor every CT examination. Thus, adrenal assessment often involves performing dedicated adrenal CT at another time. A complete washout CT evaluation for a hyperattenuating mass would involve two or three scans obtained through the adrenal glands, one scan obtained before administration of contrast material with a radiologist checking the attenuation of the gland (if a scan has not been obtained before), and two additional scans obtained after injection of contrast material. This brings up the small but still-present risks of radiation from two additional scans and potential contrast material reaction. On the other hand, chemical shift MR imaging involves no ionizing radiation, does not require radiologist monitoring, and can be performed rapidly without contrast material injection.
The role of MR imaging is thought to be limited for hyperattenuating adrenal masses because of prior studies that have shown a strong correlation between the degree of signal intensity (SI) decrease on out-of-phase MR images and CT attenuation (13). To our knowledge, however, a study aimed specifically at assessing the chemical shift MR imaging features of hyperattenuating adrenal masses has not been performed. Prior studies do contain some information regarding hyperattenuating adrenal masses, but it is limited to small numbers. The purpose of the present study was to evaluate chemical shift MR imaging for the characterization of hyperattenuating adrenal masses.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Fifty-one patients that met these criteria were identified. For these 51 patients, one observer (S.G., abdominal imaging fellow with 4 years experience reading abdominal CT images) reviewed the CT images. Image review was performed on a computer workstation (Advantage Windows v4.0; GE Medical Systems, Milwaukee, Wis). The reviewer excluded patients if they had a mass smaller than 8 mm in diameter, if the mass had an attenuation of 10 HU or less, or if it contained gross fat. An 8-mm size cutoff was chosen because of the difficulty in obtaining accurate MR imaging measurements for such small masses. An adrenal mass was considered to be hyperattenuating if a region of interest that encompassed the whole mass on unenhanced CT scans through the largest diameter of the mass contained attenuation of more than 10 HU.
Three masses were excluded because of small size, and four because of gross fat content. In eight patients, a final diagnosis could not be assigned because of inadequate follow-up, which left 36 patients (mean age, 61 years; age range, 31–77 years) with 38 masses for evaluation. One patient had one metastasis in each adrenal gland, and another patient had one adenoma in each adrenal gland. Nineteen patients were men (mean age, 64 years; age range, 33–77 years), and 17 were women (mean age, 56 years; age range, 30–67 years). The women were significantly younger than the men (P = .03). All results were examined to see if they were affected by either patient sex or age. Neither variable was significant in any analysis (P > .25 in all cases). Twenty of the 36 (56%) patients had a known primary malignancy. Approval for this study was obtained from our institutional research ethics board. Informed consent was waived by the ethics committee.
Imaging
CT was performed by using a multi–detector row helical CT scanner (Lightspeed; GE Medical Systems) without the use of intravenous contrast material. Imaging parameters included section collimation of 2.5 mm, high-quality mode, table speed of 7.5 mm/sec, 50% overlap reconstruction, 120 kV, and 280–380 mA. A single–detector row helical CT scanner (CTi; GE Medical Systems) was also used, with 3–5-mm section collimation, 50% overlap reconstruction, 120 kV, 220–300 mA, and pitch of 1:1. Multi–detector row helical CT was used in 22 of 36 patients, and single–detector row helical CT was used in 14 of 36 patients.
MR imaging was performed with a 1.5-T system (Echospeed LX; GE Medical Systems). In-phase and opposed-phase MR images were obtained by using the following parameters in the transverse and coronal planes in two consecutive breath holds with a two-dimensional spoiled gradient-echo sequence: (repetition time msec/echo time msec) 80–200/4.2–4.6, 2.1–2.3; bandwidth, 31.25–62.5 kHz; one signal acquired; section thickness, 5–7 mm; gap, 0 mm; field of view to cover the adrenals, 24–38 cm; matrix, 256 x 160–192; flip angle, 75°–90°. Dual-echo acquisition, where both the in-phase and opposed-phase MR images were obtained in the same breath hold, was used in 21 of 36 (58%) patients, while non–dual-echo acquisition was used in 15 of 36 (42%) patients.
MR Image Interpretation
One reader with 8 years of experience in reading abdominal MR images (M.A.H.) performed all quantitative MR imaging measurements (this reader was different from the one who performed the CT measurements). This reader was blinded to the CT attenuation measurements. Regions of interest were placed on each adrenal mass by using the largest ellipse possible while being careful to avoid the edges where chemical shift artifact was present. Identical regions of interest were drawn on both in-phase and opposed-phase MR images by using the copy and paste functions of the workstation. A similar method was used to measure the SI of the kidney by avoiding renal hilar fat and including the same amount of cortex and medulla. The kidney was used instead of the spleen because some of our patients underwent splenectomy, and others had hemosiderosis, which limited our ability to use the spleen as a control for SI in all cases.
The percentage SI decrease was calculated by using the following formula, which normalizes SI to renal parenchyma:
|
|
Size was defined as the largest cross-sectional diameter of the mass.
Reference Standard
The following criteria were used as a reference standard to classify all masses when pathologic findings from biopsy or surgery were not available. A mass was classified as adenoma if there was no growth during at least a 24-week period on images or if there was at least 60% enhancement washout at CT (9). There is precedent in the literature for using 6-month follow-up imaging as a reference standard for adenoma (8–10,13,14). Two patients in the present study had follow-up between 24 weeks and 6 months, one 14 days less than 6 months and the other 4 days less than 6 months. The reference standard used for diagnosis was follow-up for 33 masses, pathologic findings for four (two adenomas, one adrenocortical oncocytoma, one pheochromocytoma), and washout at CT for one. Follow-up times were as follows: for adenomas, mean of 77 weeks and range of 24–184 weeks; for metastases, mean of 17 weeks and range of 5–56 weeks. Adjusting for the follow-up time in the statistical analysis did not affect the results.
A mass was diagnosed as metastasis if it grew in a patient with a known primary malignancy. Twenty-seven of the masses were adenomas, nine were metastases, one was an adrenocortical oncocytoma, and one was a pheochromocytoma. In the nine cases of metastases, the primary tumor sites were lung in eight (one of which was a pulmonary carcinoid) and the cervix in one (a carcinoma). The mean diameter of the nonadenomas was 2.6 cm (range, 1.4–5.2 cm), and that of the adenomas was 2.2 cm (range, 1.0–5.2 cm). The sizes were not significantly different (P > .4). The threshold value for percentage SI decrease that was diagnostic of adenoma was chosen retrospectively after review of the results to maximize sensitivity while maintaining 100% specificity.
Statistical Analysis
Means were compared by using the Student t test. Exact 95% CIs were calculated for proportions by using binomial distribution. Pearson correlation coefficients were used when assessing CT attenuation versus percentage SI decrease. Age, sex, and follow-up time were each tested for their effect on mean SI decrease by using linear regression, on sensitivity by using logistic regression, and on the relationship between CT attenuation and mean SI decrease by using partial correlations. SAS version 8.2 (SAS, Cary, NC) and SPLUS-2000 (MathSoft, Seattle, Wash) statistical software was used.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
By using the spleen as a reference standard, Outwater et al (13) were able to classify three of the 10 adenomas, and all three of these were 10–20 HU. In the present study, we demonstrate a much higher sensitivity of 89% for adenoma with attenuation of 10–30 HU and 100% for adenoma with attenuation of 10–20 HU with a maintained specificity of 100%. This discrepancy may be related in part to the use of older MR imaging technology that did not allow for dual-echo single–breath-hold acquisition, which was performed in 58% of our population. This pulse sequence is not affected by misregistration between in-phase and out-of-phase sequences; thus, one would expect the comparative SI measurements to be more reliable. Unfortunately, there were insufficient data in our study to perform a comparison between dual-echo and non–dual-echo chemical shift MR imaging sequences.
Limitations
The present study has some limitations because it is retrospective. There was a potential for sample bias. The criterion of 20% SI decrease was determined retrospectively. In a prospective study, this number would have to be validated as an optimal threshold value. The fact that chemical shift MR imaging and dual-echo chemical shift MR imaging were used in the study population necessitated normalization of SI to avoid issues of SI increase between pulse sequences, which confounds the measurements of percentage SI decrease. Since dual-echo sequences do not require normalization, simpler calculations without normalization could be used in future studies that exclusively use dual-echo chemical shift MR imaging (16). To confirm this, a larger study in which dual-echo chemical shift MR imaging was used exclusively would be required.
Another limitation is that pathologic findings were used as a reference standard in a small percentage of patients; however, because of the ability of CT and MR imaging to definitively characterize adenomas, it is rare that patients with lesions diagnosed as benign will undergo surgery or biopsy in clinical practice. Although we had a small number of metastases in this study, our results concur with those in another study (16), in which nine metastases were evaluated by using a dual-echo breath-hold gradient-echo MR sequence, with all metastases having less than 20% SI decrease.
The minimum follow-up time in our study for the diagnosis of adenoma was 24 weeks, and it is possible that some adrenal metastases may have long doubling times. Some authors (7,17) have used 1-year follow-up as a lower limit; however, many leading authors (8–10,13,14) have used 6-month follow-up as a reference standard in the past. Two patients in our study had follow-up between 24 weeks and 6 months, one 14 days less than 6 months and the other 4 days less than 6 months. It is unlikely that imaging these patients at 6 months would have changed the results of our study.
Our specificity was 100%, which should be viewed with skepticism. There were no collision lesions (combined metastases and adenoma), adrenocortical cancers, or renal cell carcinoma metastases in our study group. It is in these groups where one would expect our specificity to be reduced. There is a paucity of data on the relative values of chemical shift MR imaging and washout CT for adrenocortical carcinoma. Adrenocortical cancers may exhibit SI decrease at chemical shift MR imaging (18) and may exhibit rapid washout at CT (9); however, most are large heterogeneous masses (19,20). Such large heterogeneous lesions would be sampled for biopsy or resected regardless of the other chemical shift MR imaging or CT washout findings; thus, we feel this is not a major study limitation. Similarly, the history of renal cell carcinoma would lead to consideration of biopsy regardless of imaging findings, since washout CT may not be accurate in this setting (9). Collision lesions are rare. Inhomogeneity at CT would lead to consideration of biopsy in most of these cases, as well.
Chemical shift MR imaging has some limitations as a modality. It is difficult to obtain images of adequate quality with a section thickness of less than 5 mm in a single breath hold with two-dimensional pulse sequences. Volume averaging with the SI artifact on out-of-phase MR images along the adrenal margins has the potential to artificially reduce SI, thus giving spurious high measurements of SI decrease. To avoid this, we excluded all nodules smaller than 8 mm and were careful when performing measurements to avoid adrenal margins. No nodule in our series was smaller than 10 mm. Three-dimensional acquisitions have the potential to improve the spatial resolution of MR imaging in this setting, and further studies are necessary to see if this will help in the characterization of small adrenal masses.
MR imaging is thought of as a costly modality. Prior studies have suggested that CT is more cost-effective than MR imaging; however, investigators in these studies base costs on remuneration and consider unenhanced CT alone (21). It should be kept in mind that the newer dual-echo gradient-echo MR imaging sequences allow evaluation of the adrenal gland in a single breath hold. Adrenal MR imaging can easily be completed in a 15-minute unmonitored time slot without the use of contrast material or radiologist supervision. Thus, the cost argument against MR imaging may not be as valid once one factors in the cost of intravenous contrast material and radiologist monitoring for CT and the relatively small amount of MR imaging time needed.
Clinical Implications
There is little doubt that washout CT has excellent diagnostic performance for hyperattenuating adrenal adenoma and that this is maintained at all attenuation values with sensitivities of 98% and specificities of 92%–100% (9,10). In addition, we have demonstrated that lipid-poor adenomas with attenuation higher than 30 HU at unenhanced CT are not well diagnosed with chemical shift MR imaging, and these constitute a small but notable proportion of adenomas in our study. For these reasons, washout CT will remain the modality of choice for most patients with hyperattenuating adrenal masses (11,12,22).
Although the sensitivities with chemical shift MR imaging were relatively high in our study, they are still lower than those at washout CT. Given the similar specificities of chemical shift MR imaging and washout CT, one must ask whether the decreased sensitivity of chemical shift MR imaging negates its use for assessment of all hyperattenuating adrenal masses. The risks of CT are those related to radiation dose and contrast material reaction. The risks become more of a consideration in patients with a true adrenal "incidentaloma" (those without a known primary malignancy), as they often do not have a serious comorbid condition. Such patients would have a higher prevalence of adenoma than metastases and have a high likelihood of definitive diagnosis with MR imaging. True incidentalomas are being seen more often as the indications for unenhanced CT expand (ie, appendicitis, urolithiasis, colonography, and general screening). One would expect the use of CT to continue to increase because it is driven by new technology, such as multi–detector row helical CT. As a result, in the radiology community there has been an increased awareness of the risks of radiation dose, with particular concern about the accumulated dose patients may receive during a lifetime. In this context, chemical shift MR imaging may be a reasonable alternative to washout CT to characterize a hyperdense adrenal nodule, provided it had attenuation lower than 30 HU at unenhanced CT.
In conclusion, chemical shift MR imaging has a high sensitivity for hyperattenuating adrenal adenoma with attenuation values of 10–30 HU at unenhanced CT. In certain circumstances, chemical shift MR imaging is a reasonable second imaging test for further characterization of a hyperattenuating adrenal mass.
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, M.A.H; study concepts, M.A.H.; study design, M.A.H., S.G.; literature research, S.G., K.J., M.A.H.; clinical studies, S.G., K.J., M.A.H.; data acquisition and analysis/interpretation, S.G., K.J., M.A.H.; statistical analysis, M.A.H., G.L.; manuscript preparation, M.A.H.; manuscript definition of intellectual content, M.A.H., S.G., K.J.; manuscript editing, M.A.H.; manuscript revision/review, all authors; manuscript final version approval, M.A.H., S.G., K.J.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  L. M. Ho, E. K. Paulson, M. J. Brady, T. Z. Wong, and S. T. Schindera Lipid-Poor Adenomas on Unenhanced CT: Does Histogram Analysis Increase Sensitivity Compared with a Mean Attenuation Threshold? Am. J. Roentgenol., July 1, 2008; 191(1): 234 - 238. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. T. Schindera, B. J. Soher, D. M. Delong, B. M. Dale, and E. M. Merkle Effect of Echo Time Pair Selection on Quantitative Analysis for Adrenal Tumor Characterization with In-Phase and Opposed-Phase MR Imaging: Initial Experience Radiology, July 1, 2008; 248(1): 140 - 147. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. F. Faria, S. M. Goldman, J. Szejnfeld, H. Melo, C. Kater, P. Kenney, M. P. Huayllas, G. Demarchi, V. V. Francisco, C. Andreoni, et al. Adrenal Masses: Characterization with in Vivo Proton MR Spectroscopy Initial Experience Radiology, December 1, 2007; 245(3): 788 - 797. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. K. Park, C. K. Kim, B. Kim, and J. H. Lee Comparison of Delayed Enhanced CT and Chemical Shift MR for Evaluating Hyperattenuating Incidental Adrenal Masses Radiology, June 1, 2007; 243(3): 760 - 765. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. S. Jhaveri, F. Wong, S. Ghai, and M. A. Haider Comparison of CT histogram analysis and chemical shift MRI in the characterization of indeterminate adrenal nodules. Am. J. Roentgenol., November 1, 2006; 187(5): 1303 - 1308. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Chong, K. S. Lee, H. Y. Kim, Y. K. Kim, B.-T. Kim, M. J. Chung, C. A Yi, and G. Y. Kwon Integrated PET-CT for the Characterization of Adrenal Gland Lesions in Cancer Patients: Diagnostic Efficacy and Interpretation Pitfalls RadioGraphics, November 1, 2006; 26(6): 1811 - 1824. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. H. Kim, J. M. Lee, J. K. Han, J. Y. Lee, K. H. Lee, C. J. Han, J. Y. Jo, N.-J. Yi, K.-S. Suh, K.-S. Shin, et al. Hepatic Macrosteatosis: Predicting Appropriateness of Liver Donation by Using MR Imaging--Correlation with Histopathologic Findings Radiology, July 1, 2006; 240(1): 116 - 129. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. K. Kim, S. H. Kim, Y. J. Jang, H. Ahn, C.-S. Kim, H. Park, J. W. Lee, S. Kim, and K.-S. Cho Renal Angiomyolipoma with Minimal Fat: Differentiation from Other Neoplasms at Double-Echo Chemical Shift FLASH MR Imaging Radiology, April 1, 2006; 239(1): 174 - 180. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. A. Blake, J. M. A. Slattery, M. K. Kalra, E. F. Halpern, A. J. Fischman, P. R. Mueller, and G. W. Boland Adrenal Lesions: Characterization with Fused PET/CT Image in Patients with Proved or Suspected Malignancy--Initial Experience Radiology, March 1, 2006; 238(3): 970 - 977. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Metser, E. Miller, H. Lerman, G. Lievshitz, S. Avital, and E. Even-Sapir 18F-FDG PET/CT in the Evaluation of Adrenal Masses J. Nucl. Med., January 1, 2006; 47(1): 32 - 37. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Savci, Z. Yazici, N. Sahin, S. Akgoz, and E. Tuncel Value of Chemical Shift Subtraction MRI in Characterization of Adrenal Masses Am. J. Roentgenol., January 1, 2006; 186(1): 130 - 135. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. A. Blake, M. K. Kalra, A. T. Sweeney, B. C. Lucey, M. M. Maher, D. V. Sahani, E. F. Halpern, P. R. Mueller, P. F. Hahn, and G. W. Boland Distinguishing Benign from Malignant Adrenal Masses: Multi-Detector Row CT Protocol with 10-Minute Delay Radiology, December 21, 2005; (2005) 2382041514. [Abstract] [Full Text]
|
|
|
|
|
|
J. M. Pereira, C. B. Sirlin, P. S. Pinto, and G. Casola CT and MR Imaging of Extrahepatic Fatty Masses of the Abdomen and Pelvis: Techniques, Diagnosis, Differential Diagnosis, and Pitfalls RadioGraphics, January 1, 2005; 25(1): 69 - 85. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
) and nonadenomas (
) with 99.9% CIs (error bars). Note that all metastases have an SI decrease less than 20% (dotted line).
