| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Genitourinary Imaging |
1 From the Departments of Radiology (F.C., K.H., O.H., J.F.M.) and Urology (Y.C.), Hôpital Necker, Paris, France; Department of Urology, Hôpital Cochin, Paris, France (T.F., A.V.); and Department of Radiotherapy, Institut Pierre et Marie Curie, Paris, France (L.C.). Received June 6, 2001; revision requested June 27; revision received October 1; accepted October 25. Address correspondence to F.C., 19 Avenue de Tourville, 75007 Paris, France (e-mail: frcornud@wanadoo.fr).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: At multivariate analysis, five preoperative clinical parameters—prostate-specific antigen (PSA) level, digital rectal examination (DRE) result, Gleason score and number of involved sextants at transrectal US–guided biopsy, and endorectal MR imaging result—were used to predict pT3 tumor in 336 patients who underwent radical prostatectomy. On the basis of results of the first four examinations, multivariate analysis was performed also to determine predictors of a positive MR imaging study.
RESULTS: Significant predictors of pT3 tumor were positive MR imaging result (P < 2 x 10-8), more than one sextant involved at biopsy (P < 5 x 10-5), and PSA level greater than 10 ng/mL (P < 7 x 10-3). Significant predictors of a positive MR imaging result were three or more sextants involved at biopsy (P < 10-5), positive DRE result (P < 5 x 10-3), and PSA level greater than 10 ng/mL (P < 16 x 10-3). In the subgroup of 175 patients who had at least three positive biopsy specimens, the sensitivity of MR imaging was 50% for detection of occult pT3 tumor and 69% for detection of extensive pT3 tumor. The overall specificity of MR imaging was 95%.
CONCLUSION: Endorectal MR imaging seems to be indicated in carefully selected patients—specifically, those with three or more positive biopsy specimens, a palpable tumor, and/or a PSA level greater than 10 ng/mL.
© RSNA, 2002
Index terms: Prostate, biopsy, 844.1261 • Prostate neoplasms, 844.32 • Prostate neoplasms, MR, 844.121411, 844.121412, 844.121416
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
Consequently, some authors (1) believe that imaging examinations, especially endorectal magnetic resonance (MR) imaging, are unnecessary prior to treatment. However, the results of a study performed by D’Amico et al (5) showed the imperfections of using these combined criteria in patients with intermediate risk of extraprostatic extension, or pT3 tumor, as defined on the basis of a PSA level of 11–20 ng/mL and a Gleason score of 5–7. In such patients, the percentage of pT3 tumors ranges from 18% to 53% (4); this means that the described combined-criteria approach has poor reliability in some patients. Results of the D’Amico et al study (5) showed that the addition of endorectal MR imaging increased the efficiency of pT3 tumor detection relative to detection with the PSA level and Gleason score.
Several authors (6–15) have also recommended including the percentage of positive biopsy specimens (calculated on the basis of the number of sextants invaded by tumor) in addition to the Gleason score at biopsy in the criteria for detecting extraprostatic extension. The number of involved sextants indirectly reflects tumor volume (16), which is closely linked to the risk of extraprostatic extension (17) and highly reliable for estimating the risk of extraprostatic extension, independently of the PSA level and the Gleason score. D’Amico et al (5,18,19) applied this variable to patients with intermediate risk of pT3 tumors, with endorectal MR imaging being indicated when at least 50% of the biopsy specimens were positive (ie, tumor invasion in three of six sampled sextants).
In a series of 336 patients with clinically localized prostate tumors who were to undergo radical prostatectomy, we conducted a multivariate analysis of the following available clinical parameters: DRE result, PSA level, Gleason score, number of invaded sextants at US-guided biopsy, and endorectal MR imaging result. The purpose of our study was to identify the factor(s) most predictive of pT3 tumor and the factor(s) most predictive of a positive MR imaging result.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
Sextant biopsies were performed transrectally with an 18-gauge needle mounted on a spring-loaded commercial biopsy device (Pro-Mag 2.2; Meditech, Paris, France). Directed biopsies involved two needle passes into all of the areas that were hypoechoic on transrectal US images. Additional sextant biopsies involved one ipsilateral needle pass into each isoechoic sextant adjacent to any abnormal area and one core-needle specimen taken from each contralateral sextant midway between the midline and the lateral aspect of the prostate. At least six to eight core-needle biopsy specimens were taken, but the recorded criterion used to define the number of positive biopsy specimens was the number of sextants (of six) involved by tumor. Since 1997, we have incorporated an extended sextant biopsy protocol (20) that includes twelve biopsies for patients without transrectal US or DRE abnormalities, but in this study only the number of sextants involved by tumor was taken into account to define the number of positive biopsy specimens.
Pathologic staging was performed by three uropathologists (including A.V.) according to a modified Stanford technique. Surgical margins were not marked with ink before sectioning. The prostate gland was fixed in 10% buffered formaldehyde. The seminal vesicles and apex were separated and examined on sagittal sections. The middle portion of the gland was sectioned axially at 5-mm intervals.
Pathologic T classification was used to assess intra- and extraprostatic tumor spread. Capsular penetration was classified, according to the Epstein classification system, as focal (ie, microscopic) if a few tumor cells were present outside the prostate and as established if more extensive extraprostatic spread was present (21). Seminal vesicle invasion (SVI) was defined as microscopic if it involved only the intraprostatic portion of the vesicles and macroscopic if it involved the extraprostatic portion. Surgical margins were considered to be positive when the tumor showed histologic extension to the surface. In 50 (14.9%) patients, a single margin (apical in 36 cases, posterior or posterolateral in 11 cases, anterior or anterolateral in 3 cases) that was positive for tumor was observed in an area without periprostatic tissue, with no signs of capsular penetration. These patients were classified as having pT2 tumors with a positive margin due to inadvertent incision of the capsule during surgery (22).
Endorectal MR imaging was performed 2–3 weeks after biopsy. MR imaging findings were assessed by means of consensus between two readers (F.C., K.H.), who began their training in the interpretation of endorectal MR images 2 years before the beginning of this study and were aware of the DRE, sextant biopsy, and PSA assay results. The two readers interpreted the images together, and discrepancies, regarding signs of capsular penetration, occurred in 13% (45 of 336) of cases. A third reader (O.H.) then interpreted the images, and all three readers agreed by means of consensus.
We performed endorectal MR imaging by using a 1.5-T superconducting magnet (Signa; GE Medical Systems, Milwaukee, Wis) and an endorectal surface coil (MedRad, Pittsburgh, Pa) coupled (since 1997) to an anterior surface coil. After a gradient-echo localizer sequence was performed to check the coil position, fast spin-echo MR images (3,000/102 [repetition time msec/echo time msec], 4-mm section thickness, 0.4-mm intersection gap, two signals acquired, 16-cm field of view, 512 x 256 matrix, no phase wrap) were acquired. Transverse T1-weighted MR images were acquired to detect biopsy artifacts and assess lymph node involvement. T2-weighted MR images were acquired in the transverse and coronal planes; the coronal images were focused on the caudal junction of the vas deferens and seminal vesicles.
Capsular penetration was diagnosed when MR imaging depicted irregular bulging of the prostate associated with capsular signal intensity disruption and/or periprostatic fat infiltration. The sites of capsular penetration were labeled according to the source of penetration (ie, apex, including invasion of periapical fat and/or of the striated sphincter [23], or posterior or posterolateral aspects of the prostate at the middle portion or prostate base and bladder neck) to ensure that the ECE specified in the pathology report was not found in a location that was different from that reported at MR imaging. Signs of periprostatic fat infiltration included tumor signal intensity within periprostatic fat, obliteration of the rectoprostatic angle, and/or asymmetry of the neurovascular bundles at MR imaging (24). Indirect signs of ECE, such as smooth bulging and broad tumor contact (25,26), were not included in the assessment. SVI was diagnosed when focal low signal intensity was present in one or both seminal vesicles at MR imaging. Because SVI is absent when biopsy reveals no involvement of the prostate base on both sides (27), focal thickening of the tubular walls at T2-weighted MR imaging was considered to be evidence of SVI (28) only when biopsy results showed involvement of at least the ipsilateral prostate base.
The sensitivity, specificity, and accuracy of the DRE result, PSA level (
10, 11–20, or >20 ng/mL), Gleason score at biopsy (<7, 7, or >7), and number of sextants involved at biopsy (1, 2, 3, or >3) in the prediction of pT3 tumor stage—that is, ECE and/or SVI—were assessed and compared with the corresponding values at MR imaging. Then, the sensitivity, specificity, and accuracy of MR imaging in the prediction of ECE and SVI were assessed separately according to the results of DRE, PSA assay (10, 11–20, or >20 ng/mL), and Gleason scoring at biopsy (<7, 7, or >7). We then performed multivariate analysis (ie, logistic regression) to test the five preoperative parameters as predictors of the pathologic stage and to identify those parameters that were predictive of a positive MR imaging result. Finally, on the basis of the multivariate analysis results, different combinations of the clinically significant parameters were used to assess the prevalence of pT3 tumor and the sensitivity, specificity, and accuracy of MR imaging calculated by using each parameter combination.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
|
In the subgroup of 69 patients with extensive extraprostatic spread, the number of pT3 tumors differed significantly only between the patients with at least three positive biopsy specimens (58 [33%] of 175 patients) and those who had one to two positive biopsy specimens (11 [7%] of 161 patients, P < 10-7).
Sensitivity, Specificity, and Accuracy of Preoperative Criteria for Prediction of pT3 Tumor Stage
The performances of each of the five parameters (DRE result, PSA level, Gleason score, number of positive biopsy specimens, and MR imaging result) in the prediction of pT3 tumor stage and extensive pT3 tumor stage—that is, ECE and/or SVI—are summarized in Table 2. The performance of MR imaging in the prediction of either ECE or SVI based on the results of the four other parameter examinations is summarized in Table 3.
|
|
Overall, among the 113 patients with pT3 cancer, 44 had microscopic extraprostatic spread, and none of these tumors was detected at MR imaging (0% sensitivity). Sixty-nine patients had extensive extraprostatic spread, and 43 (62%) of these tumors were detected at MR imaging. The overall sensitivity of MR imaging for detection of occult pT3 tumor was 40% (45 of 113 patients). MR imaging resulted in the erroneous diagnosis of extraprostatic spread in 12 patients, including 10 patients with a false-positive diagnosis of ECE and two with a false-positive diagnosis of both ECE and SVI, and thus had an overall specificity of 95% (211 of 223 patients).
Multivariate Analysis for Prediction of pT3 Tumor and a Positive MR Imaging Result
Significant predictors of pT3 tumor were positive MR imaging result (P < 2 x 10-8), more than one sextant involved at biopsy (P < 5 x 10-5), and PSA level greater than 10 ng/mL (P < 7 x 10-3) (Table 4). Significant predictors of a positive MR imaging result were three or more positive biopsy specimens (P < 10-5), positive DRE result (P < 5 x 10-3), and PSA level greater than 10 ng/mL (P < 16 x 10-3) (Table 4).
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
Whether it is because of extraprostatic extension (6–14), positive margins (16,29), or increasing PSA levels after prostatectomy (13,30–32), the risk of failed radical prostatectomy increases with the number of positive biopsy specimens. In patients with one or two positive biopsy specimens, the risk ranges from 7.4% to 40.0% (7,8,14–16,30,31,33). This range has been lower in studies in which the investigators selected those patients with one positive biopsy specimen, in whom the risk of prostatectomy failure does not exceed 20% (7,8,14,16,30,33), with one exception (31). In patients with three or more positive biopsy specimens, the risk of prostatectomy failure is at least 45% (6–8,11–13,15,30,32,33), again with one exception (29).
The results in patients with more than three positive biopsy specimens have been contradictory: The risk reached 93% in one study (10), but it did not increase substantially in the study of D’Amico et al (33) (55%, 45%, and 50%, respectively, in patients with four, five, and six positive biopsy specimens). Our study results confirm most of these data: The percentages of pT3 tumors were 6%, 30%, 46%, and 53% in patients with one, two, three, and more than three positive biopsy specimens, respectively. The reason there was no marked increase in the number of pT3 tumors in the patients with more than three positive biopsy specimens is unclear, although the other risk factors (ie, PSA level and Gleason score) were not more unfavorable when the number of positive biopsy specimens increased from three to more than three. This was also the case in the study of D’Amico et al (33), in which all patients with at least three positive biopsy specimens had the same intermediate risk of pT3 tumor, which was defined on the basis of a PSA level of 4–20 ng/mL and a Gleason score of 7 or lower.
Our study results also show that the number of positive biopsy specimens can be used to identify patients with a risk of extensive (ie, established) pT3 tumor (who have a poor outcome after surgery), whether it is because of extensive ECE (34) or SVI (35). In our study, the percentage of extensive pT3 tumors was 7% (11 of 161 patients) among the patients with one or two positive biopsy specimens and almost five times higher (58 [33%] of 175 patients) when the number of positive biopsy specimens reached three. These results have important implications with regard to the indications for endorectal MR imaging, which can depict only extensive pT3 tumors.
Endorectal MR imaging has been used to predict the extraprostatic extension of clinically localized tumors since 1991 (28), although many urologists do not use it because they believe it is not sensitive or specific enough to predict pT3 tumor (1). Moreover, one limitation of our study is that we did not assess MR imaging for the prediction of patient outcome after surgery and thus did not take into account the fact that some patients with pT3 tumor can be cured with surgery. However, in one report (21), patients with focal ECE showed no evidence of progression at long-term follow-up after surgery, and it has been shown that endorectal MR imaging results that are negative or positive for ECE or SVI help to separate patients with clinically confined cancer and intermediate risk of extraprostatic spread into groups with a 78% versus 21% (P < 10-4) 3-year rate of actuarial freedom from PSA failure (36). Thus, a positive MR imaging result has important implications for patient outcome after surgery.
Several authors (24,37), similar to us in our study, have reported the conditions in which MR imaging should be used—that is, to yield the best specificity. These conditions include studies performed by experienced radiologists (19), the exclusion of indirect signs of extraprostatic extension (25,26), and the use of only direct signs of capsular effraction or SVI (24,38,39). Several authors (37,40) also recommend that all other preoperative parameters be established before the interpretation of MR imaging findings, although not being blinded to DRE, biopsy, and PSA results could be considered a flaw in the interpretation of MR images. However, this is the way we use MR imaging in our clinical practice, and in these conditions, the specificity of MR imaging can reach about 95%; this reduces to an acceptable minimum the percentage of patients for whom radical prostatectomy would not be indicated because of a false-positive MR imaging result. Such specificity is obtained only at the cost of weak sensitivity (37,40), which was lower than 40% in our study, and this is a major limitation to the routine use of endorectal MR imaging for assessment of local extension of clinically localized prostate cancer. In our study, only 69 (21%) of the 336 patients were candidates for MR imaging.
At multivariate analysis, we integrated all available preoperative parameters to select candidates for MR imaging. Without MR imaging, the factor that was most predictive of pT3 tumor was number of positive biopsy specimens, followed by PSA level, as in other studies (6,9,11,13). In other studies, Gleason score was the most meaningful cofactor (10,12,14,15,41) for predicting pT3 tumor. Nevertheless, multivariate analysis results showed that MR imaging added significant predictive value to the value of the number of positive biopsy specimens. MR imaging is therefore justified when the factors that are predictive of a positive result can be determined.
Multivariate analysis results showed also that a number of positive biopsy specimens of three or more is the most powerful predictor of a positive MR imaging result. The importance of the number of positive biopsy specimens in justifying endorectal MR imaging was repeatedly emphasized by D’Amico et al (19,33) in studies involving patients with intermediate risk of pT3 tumor, which was defined on the basis of a PSA level of 4–20 ng/mL, a Gleason score of 7 or lower, and at least three positive biopsy specimens. In one study (33) involving these patients, the results showed that the likelihood of pT3 tumor increased from 45% to 100% when the MR imaging result was positive. Results of the other study (19) showed that negative MR imaging results in the same group of patients increased the likelihood of organ-confined cancer in the prostatectomy specimen from 32% to 64%. Our study results are in agreement with these results because they confirm that MR imaging is not indicated for patients with fewer than three positive biopsy specimens, given the very low prevalence of extensive pT3 tumor observed (in 11 [7%] of 161 patients).
The cutoff of three positive biopsy specimens led to the identification of 84% (58 of 69 patients) of the patients with extensive pT3 tumor and to the use of MR imaging to detect extraprostatic extension in 69% (40 of 58) of these patients. In the same way (see Results section), a negative MR imaging result significantly increased the likelihood of local tumors (ie, tumors with at most microscopic extraprostatic extension) from 66% in only the patients with three positive biopsy specimens to 85% when a positive MR imaging result was included in the analysis.
Our study results also show the very significant predictive value of an abnormal DRE result and a PSA level greater than 10 ng/mL in predicting a positive MR imaging result. This means that MR imaging is not indicated in patients with nonpalpable cancer and a PSA level lower than 10 ng/mL, even in those with at least three positive biopsy specimens, given the very low prevalence of extensive pT3 tumor (11%) that we observed in this patient group. This conclusion was also suggested in one of the studies performed by D’Amico et al (19) and was demonstrated in a study performed by Cornud et al (40). Patient selection that is based on three positive biopsy specimens, positive DRE results, and a PSA level of greater than 10 ng/mL greatly restricts the indications for MR imaging. In one of the studies performed by D’Amico et al (33), MR imaging was indicated in only 7.5% of patients. In our study population, if we considered that MR imaging cannot be indicated for a threshold prevalence of extensive pT3 tumor of less than 20%, then no patient with fewer than three positive biopsy specimens would be considered a candidate for MR imaging, regardless of the DRE or PSA assay results (Tables 5 and 6). Thus, only the patients with at least three positive biopsy specimens would be considered candidates for MR imaging. In addition, if we followed the recommendations of Jager et al (37)—namely, performing endorectal MR imaging in patients in whom the risk of extensive pT3 tumor is at least 39%—then the data in Table 7 would show that MR imaging was really indicated in only 37% (65 of 175) of the patients with at least three positive biopsy specimens—that is, 19% (65 of 336) of all the patients in our series. The data in Table 7 also show how MR imaging can still be indicated when a lower threshold prevalence of extensive pT3 tumor is acceptable. It might be valid to consider MR imaging when there is a prevalence threshold of 30%—for example, in patients with three or more positive biopsy specimens and a PSA level greater than 10 ng/mL. If a threshold of 20% were acceptable, then a criterion of three or more positive biopsy specimens and either a positive DRE result or a PSA level greater than 10 ng/mL would indicate the need for MR imaging.
In conclusion, endorectal MR imaging is a reliable way of predicting clinically occult pT3 tumor in patients with clinically localized prostate cancer. However, the very high specificity of MR imaging is offset by low sensitivity; this means that this examination is useful only in carefully selected patients with a risk of extensive pT3 tumor—that is, those with at least three positive biopsy specimens, positive DRE results, and/or a PSA level greater than 10 ng/mL.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, J.F.M., F.C.; study concepts, F.C., T.F., Y.C.; study design, F.C., O.H.; literature research, F.C.; clinical studies, F.C., K.H., O.H.; data acquisition, F.C., K.H.; data analysis/interpretation, F.C., K.H., O.H., L.C.; statistical analysis, L.C.; manuscript preparation, F.C., O.H.; manuscript definition of intellectual content, F.C., T.F., L.C., A.V.; manuscript editing, F.C., T.F.; manuscript revision/review, L.C., T.F., A.V.; manuscript final version approval, J.F.M.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. K. Kuhl, F. Traber, J. Gieseke, W. Drahanowsky, N. Morakkabati-Spitz, W. Willinek, M. von Falkenhausen, C. Manka, and H. H. Schild Whole-Body High-Field-Strength (3.0-T) MR Imaging in Clinical Practice * Part II. Technical Considerations and Clinical Applications Radiology, April 1, 2008; 247(1): 16 - 35. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. N. Bloch, E. Furman-Haran, T. H. Helbich, R. E. Lenkinski, H. Degani, C. Kratzik, M. Susani, A. Haitel, S. Jaromi, L. Ngo, et al. Prostate Cancer: Accurate Determination of Extracapsular Extension with High-Spatial-Resolution Dynamic Contrast-enhanced and T2-weighted MR Imaging Initial Results Radiology, October 1, 2007; 245(1): 176 - 185. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. W. T. P. J. Heijmink, J. J. Futterer, T. Hambrock, S. Takahashi, T. W. J. Scheenen, H. J. Huisman, C. A. Hulsbergen-Van de Kaa, B. C. Knipscheer, L. A. L. M. Kiemeney, J. A. Witjes, et al. Prostate Cancer: Body-Array versus Endorectal Coil MR Imaging at 3 T--Comparison of Image Quality, Localization, and Staging Performance Radiology, July 1, 2007; 244(1): 184 - 195. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Hricak, P. L. Choyke, S. C. Eberhardt, S. A. Leibel, and P. T. Scardino Imaging Prostate Cancer: A Multidisciplinary Perspective Radiology, April 1, 2007; 243(1): 28 - 53. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Wang, J. Zhang, L. H. Schwartz, H. Eisenberg, N. M. Ishill, C. S. Moskowitz, P. Scardino, and H. Hricak Incremental Value of Multiplanar Cross-Referencing for Prostate Cancer Staging with Endorectal MRI Am. J. Roentgenol., January 1, 2007; 188(1): 99 - 104. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Wang, H. Hricak, M. W. Kattan, H. N. Chen, K. Kuroiwa, H. F. Eisenberg, and P. T. Scardino Prediction of Seminal Vesicle Invasion in Prostate Cancer: Incremental Value of Adding Endorectal MR Imaging to the Kattan Nomogram Radiology, December 1, 2006; 242(1): 182 - 188. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Wang, M. Mullerad, H.-N. Chen, S. C. Eberhardt, M. W. Kattan, P. T. Scardino, and H. Hricak Prostate Cancer: Incremental Value of Endorectal MR Imaging Findings for Prediction of Extracapsular Extension Radiology, July 1, 2004; 232(1): 133 - 139. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Dhingsa, A. Qayyum, F. V. Coakley, Y. Lu, K. D. Jones, M. G. Swanson, P. R. Carroll, H. Hricak, and J. Kurhanewicz Prostate Cancer Localization with Endorectal MR Imaging and MR Spectroscopic Imaging: Effect of Clinical Data on Reader Accuracy Radiology, January 1, 2004; 230(1): 215 - 220. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. L. Zakian, S. Eberhardt, H. Hricak, A. Shukla-Dave, S. Kleinman, M. Muruganandham, K. Sircar, M. W. Kattan, V. E. Reuter, P. T. Scardino, et al. Transition Zone Prostate Cancer: Metabolic Characteristics at H MR Spectroscopic Imaging--Initial Results Radiology, October 1, 2003; 229(1): 241 - 247. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
