AJR ARRS Membership
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wang, L.
Right arrow Articles by Hricak, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wang, L.
Right arrow Articles by Hricak, H.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?
DOI:10.2214/AJR.05.1783
AJR 2007; 188:99-104
© American Roentgen Ray Society

Original Research

Incremental Value of Multiplanar Cross-Referencing for Prostate Cancer Staging with Endorectal MRI

Liang Wang1, Jingbo Zhang1, Lawrence H. Schwartz1, Halley Eisenberg1, Nicole M. Ishill2, Chaya S. Moskowitz2, Peter Scardino3 and Hedvig Hricak1

1 Department of Radiology, Memorial Sloan-Kettering Cancer Center, 1275 York Ave., Rm. C-278, New York, NY 10021.
2 Department of Epidemiology and Biostatistics, Memorial Sloan-Kettering Cancer Center, New York, NY 10021.
3 Department of Urology, Memorial Sloan-Kettering Cancer Center, New York, NY 10021.

Received October 10, 2005; accepted after revision December 16, 2005.

 
Supported by National Institutes of Health grant NIH R01 CA76423.

Address correspondence to H. Hricak.


Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
OBJECTIVE. The purpose of this study was to assess whether use of the PACS cross-referencing tool in 3D MRI improves tumor staging of prostate cancer when pathologic findings are used as the reference standard.

MATERIALS AND METHODS. The institutional review board granted a waiver of informed consent for the study. Endorectal MRI at 1.5 T was performed before radical prostatectomy in 255 consecutive patients. Two radiologists unaware of the clinical data retrospectively and independently interpreted MR images without and with cross-referencing to predict the presence of extracapsular extension (ECE) and seminal vesicle invasion (SVI). Histopathologic findings were used as the reference standard. Area under the receiver operating characteristics curve (AUC), sensitivity and specificity, and weighted kappa statistics were calculated.

RESULTS. At histologic examination, 68 (27%) of the patients were found to have ECE and 13 (5%) of the patients to have SVI; the latter all had ECE. In detecting ECE, both reviewers had a higher AUC using cross-referencing (p < 0.001 for both). The weighted kappa value was 0.56 for MRI alone and 0.76 for MRI with cross-referencing, indicating fair to good interobserver agreement. Sensitivity and specificity for ECE with MRI alone and with cross-referencing were 43% and 94% and 57% and 100% for reviewer 1 and 40% and 93% and 59% and 98% for reviewer 2, respectively. In detecting SVI, both reviewers had a higher AUC with cross-referencing (p = 0.007 and p = 0.056 for reviewers 1 and 2, respectively). Reviewer 1 benefited much more from cross-referencing than did reviewer 2. The weighted kappa statistic was 0.69 for MRI alone and the same with cross-referencing, indicating good interobserver agreement. Sensitivity and specificity for SVI with MRI alone and with cross-referencing, respectively, were 23% and 83% and 46% and 93% for reviewer 1 and 31% and 91% and 54% and 95% for reviewer 2.

CONCLUSION. PACS cross-referencing significantly improves tumor staging of prostate cancer with 3D MRI. Some reviewers benefit more than others from use of this tool.

Keywords: digital imaging • MR coils • MRI • neoplasms • oncologic imaging • PACS • prostate • tumor staging


Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Prostate cancer affects men of all races, cultures, and ethnic backgrounds and is a major public health and economic burden in the United States and other industrialized countries [1, 2]. In the past decade, there has been a dramatic downward trend in the stage of prostate cancer determined at diagnosis [1, 3]. Preoperative identification of extracapsular extension (ECE) and seminal vesicle invasion (SVI) is an important factor in staging and prognosis that may modify treatment decisions and treatment planning [4, 5]. Because it improves surgical planning, identification of these characteristics increases the chances that the tumor will be resected completely with only minimal damage to the surrounding tissue so important to recovery of normal function [6, 7].

Probability of the presence of ECE and SVI currently is best determined from staging nomograms with which pathologic stage is estimated from the pretreatment level of prostate-specific antigen, clinical stage, and Gleason grade in the biopsy specimen [7-9]. However, staging nomograms are limited because they do not incorporate the results of imaging studies that could assist in prediction of ECE and SVI [7-9].

Because of high spatial resolution, superior contrast resolution, multiplanar capability, and large field of view, MRI has played an increasingly important role in prostate cancer staging. However, although it has high specificity (ECE specificity, 47-99%; SVI specificity, 81-99%) [10-14], MRI has widely varying sensitivity (ECE sensitivity, 22-80%; SVI sensitivity, 34-71%) [15-17]. Many of the difficulties in prostate cancer staging have been caused by poor understanding of the anatomic features of the prostate owing to its small size, presence of periprostatic structures [18, 19], and the partial volume effect on MRI [20]. The partial volume effect arises in volumetric images when more than one tissue type is present in a voxel. In such cases, voxel intensity depends not only on the imaging sequence and tissue properties but also on the proportion of each tissue type present in the voxel [20]. Depending on the direction and thickness of the MRI slice, the border between the tissues can become unclear. Presence of the partial volume effect can lead to false assessment of tumor capsules and tumor invasion [21].


Figure 1
View larger version (153K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1A —53-year-old man with clinical stage T1c prostate carcinoma with Gleason score of 3 + 3 and prostate-specific antigen level of 11.9 ng/mL. MRI without PACS cross-referencing indicated possible extracapsular extension (ECE) at right base and possible right seminal vesicle invasion (SVI) (scores of 3 and 3, respectively). However, MRI with PACS cross-referencing indicated definite ECE and definite SVI (scores of 5 and 5, respectively). Axial T2-weighted image shows point of interest (crosshairs) selected by reviewer in low-signal-intensity area using cross-referencing tool. Irregular nodularity is seen in right seminal vesicle (arrowhead), indicating seminal vesicle invasion.

 


Figure 2
View larger version (168K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1B —53-year-old man with clinical stage T1c prostate carcinoma with Gleason score of 3 + 3 and prostate-specific antigen level of 11.9 ng/mL. MRI without PACS cross-referencing indicated possible extracapsular extension (ECE) at right base and possible right seminal vesicle invasion (SVI) (scores of 3 and 3, respectively). However, MRI with PACS cross-referencing indicated definite ECE and definite SVI (scores of 5 and 5, respectively). T2-weighted coronal MR image with cross-referencing shows ECE (arrow) at right base and right seminal vesicle invasion (arrowheads). Crosshairs indicate point of interest identified by cross-referencing tool. Relationship between seminal vesicle and central zone is shown clearly.

 


Figure 3
View larger version (172K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1C —53-year-old man with clinical stage T1c prostate carcinoma with Gleason score of 3 + 3 and prostate-specific antigen level of 11.9 ng/mL. MRI without PACS cross-referencing indicated possible extracapsular extension (ECE) at right base and possible right seminal vesicle invasion (SVI) (scores of 3 and 3, respectively). However, MRI with PACS cross-referencing indicated definite ECE and definite SVI (scores of 5 and 5, respectively). Sagittal T2-weighted MR image shows right seminal vesicle invasion (arrowheads). Crosshairs indicate point of interest identified by cross-referencing tool.

 


Figure 4
View larger version (53K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1D —53-year-old man with clinical stage T1c prostate carcinoma with Gleason score of 3 + 3 and prostate-specific antigen level of 11.9 ng/mL. MRI without PACS cross-referencing indicated possible extracapsular extension (ECE) at right base and possible right seminal vesicle invasion (SVI) (scores of 3 and 3, respectively). However, MRI with PACS cross-referencing indicated definite ECE and definite SVI (scores of 5 and 5, respectively). Histologic photograph shows whole-mount sections that confirm presence of ECE (red) at right base and right seminal vesicle invasion (blue). RSV = right seminal vesicle.

 
PACS technology facilitates the display and distribution of digital images [22]. With a PACS, medical images obtained with various techniques, such as CT, MRI, sonography, and digital projection radiography [23], and other pertinent information are transmitted to various, sometimes remote, locations over networks. Display of the images on computer workstations for soft-copy viewing allows simultaneous consultations and almost instant reporting from radiologists working in several locations [23]. The PACS workstation (Centricity RA 1000, GE Healthcare) at our institution has a cross-referencing feature whereby selection of a voxel in any one plane highlights the corresponding voxel in the intersecting planes. This feature is also available on systems such as the Synapse PACS (Fujifilm) and iSite PACS (Philips Medical Systems). Although it may be tempting to assume that new technology will enhance radiologists' performance, this is not always the case. For example, current literature suggests that accuracy of interpretation of sonograms is similar for images interpreted on a monitor and those interpreted on film [24]. The purpose of our study was to assess whether use of PACS cross-referencing improves tumor staging of prostate cancer with 3D MRI when pathologic findings are the reference standard.


Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Patient Characteristics
The institutional review board granted exempt status for this retrospective, single-institution cross-sectional study with a waiver of informed consent. Patient data were collected and handled in accordance with Health Insurance Portability and Accountability Act regulations. From March 22, 2004, through January 7, 2005, 255 patients with prostate cancer were consecutively referred from the urology department to undergo endorectal MRI before radical retropubic prostatectomy performed by one of six attending surgeons at our institution. Mean patient age was 59 years (range, 40-86 years). None of the patients received neoadjuvant hormonal or radiation therapy before radical prostatectomy. All patients had a tissue diagnosis of prostate cancer based on biopsy results. Clinical serum level of prostate-specific antigen, Gleason grade at biopsy, clinical stage, greatest percentage of cancer in all biopsy cores, percentage of cores with positive results in all biopsy cores, presence of perineural invasion, and MR data were recorded retrospectively from medical records.

MRI Technique
Endorectal MRI was performed with a 1.5-T whole-body MRI system (Signa Horizon, GE Healthcare). Examinations were performed with the patients in the supine position, a body coil for excitation, and a pelvic phased-array coil (GE Healthcare) in combination with a commercially available balloon-covered expandable endorectal coil (Medrad) for signal reception. T1-weighted axial and spin-echo images were obtained from the aortic bifurcation to the symphysis pubis with the following parameters: TR/TE, 700/8; slice thickness, 5 mm; interslice gap, 1 mm; field of view, 24 cm; matrix size, 256 x 192; frequency direction, transverse (to prevent obstruction of the pelvic node by endorectal coil motion artifact); and number of excitations, one. Thin-section, high-spatial-resolution axial and coronal T2-weighted fast spin-echo images of the prostate and seminal vesicles were obtained with the following parameters: TR/effective TE, 5,000/96; echo-train length, 16; slice thickness, 3 mm; interslice gap, 0 mm; field of view, 14 cm; matrix size, 256 x 192; frequency direction, anteroposterior (to prevent obstruction of the prostate by endorectal coil motion artifact); and number of excitations, 4. All MRI data were entered into the PACS of our radiology department.


Figure 5
View larger version (136K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2A —63-year-old man with clinical stage T1c prostate carcinoma with Gleason score of 4 + 4, prostate-specific antigen level of 12.9 ng/mL. MRI without PACS cross-referencing indicated possible extracapsular extension (ECE) at left base and possible left seminal vesicle invasion (SVI) (scores of 3 and 3, respectively). However, cross-referenced MR images (A-C) indicated no ECE and no SVI (scores of 1 and 1, respectively). Axial T2-weighted MR image shows point of interest (crosshairs) identified by cross-referencing tool.

 


Figure 6
View larger version (156K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2B —63-year-old man with clinical stage T1c prostate carcinoma with Gleason score of 4 + 4, prostate-specific antigen level of 12.9 ng/mL. MRI without PACS cross-referencing indicated possible extracapsular extension (ECE) at left base and possible left seminal vesicle invasion (SVI) (scores of 3 and 3, respectively). However, cross-referenced MR images (A-C) indicated no ECE and no SVI (scores of 1 and 1, respectively). Coronal T2-weighted MR image shows interface (arrow) between seminal vesicle and central zone of prostate. Crosshairs indicate point of interest identified by cross-referencing tool.

 


Figure 7
View larger version (161K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2C —63-year-old man with clinical stage T1c prostate carcinoma with Gleason score of 4 + 4, prostate-specific antigen level of 12.9 ng/mL. MRI without PACS cross-referencing indicated possible extracapsular extension (ECE) at left base and possible left seminal vesicle invasion (SVI) (scores of 3 and 3, respectively). However, cross-referenced MR images (A-C) indicated no ECE and no SVI (scores of 1 and 1, respectively). Sagittal T2-weighted image with cross-referencing shows point of interest selected by reviewer (crosshairs).

 


Figure 8
View larger version (57K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2D —63-year-old man with clinical stage T1c prostate carcinoma with Gleason score of 4 + 4, prostate-specific antigen level of 12.9 ng/mL. MRI without PACS cross-referencing indicated possible extracapsular extension (ECE) at left base and possible left seminal vesicle invasion (SVI) (scores of 3 and 3, respectively). However, cross-referenced MR images (A-C) indicated no ECE and no SVI (scores of 1 and 1, respectively). Histopathologic photograph shows whole-mount sections confirming absence of extracapsular extension and seminal vesicle invasion at left base. RSV = right seminal vesicle, LSV = left seminal vesicle.

 
MRI Analysis on PACS Workstation
MR images were retrospectively analyzed independently by two radiologists who were aware of the hypothesis being tested but unaware of the clinical, surgical, and histologic findings for the 255 patients who had undergone radical prostatectomy. Reviewer 1 had 4 years of research experience in MRI of the prostate during which he had interpreted more than 1,000 prostate MR images. Reviewer 2 had completed a dedicated body-MRI fellowship and had interpreted more than 500 prostate MR images, most in a 1-year span. Criteria for signs of ECE on MR images included capsular irregularity, bulging of the capsule, capsular retraction, obliteration of the rectoprostatic angle, and asymmetry or direct involvement of the neurovascular bundles [10, 25]. Criteria for SVI were a focal low-signal-intensity mass or diffuse enlargement with low signal intensity and loss of the perceptible vesical wall on both T1- and T2-weighted sequences [25].

At the PACS workstation, the reviewers opened the examinations in the axial, coronal, and sagittal image series and assigned separate scores for the probabilities of ECE and SVI on a scale of 1-5 (1, definitely absent; 2, probably absent; 3, possibly present; 4, probably present; 5, definitely present). These scores were designated MRI without PACS cross-referencing. The reviewers then activated the cross-referencing feature. With this feature, when a reviewer selects a point of interest (or voxel) in any one plane (i.e., image series), an image in each intersecting plane is automatically displayed with a target symbol (or crosshairs) on the corresponding voxel. After using the cross-referencing tool, the radiologists assigned scores for ECE and SVI, again using the 5-point scale. These scores were designated MRI with PACS cross-referencing (Figs. 1A, 1B, 1C, 1D, 2A, 2B, 2C, and 2D).

Pathologic Assessment and Comparison with MR Images
Core biopsy specimens were evaluated for Gleason grade, greatest percentage of cancer in all biopsy cores, percentage of cores with positive findings in all biopsy cores, and presence of perineural invasion. All prostatectomy specimens were inked with India ink tattoo dye (green dye on right, blue dye on left) and fixed in 10% formalin for 36 hours. The distal 5 mm of the apex was amputated and coned. The rest of the gland was serially sectioned from apex to base for acquisition of axial slices at 3-mm intervals (axial step sections) and submitted in entirety for paraffin embedding as whole mounts. The seminal vesicles were amputated and submitted separately. After paraffin embedding, microsections were placed on glass slides and stained with H and E. The axial pathologic sections were numbered consecutively from apex to base, and the cancerous areas were mapped in each section with marker. Pathologic stage and surgical Gleason score were determined for each patient. ECE was defined as invasion of prostate cancer beyond the prostate capsule into the periprostatic soft tissue. SVI was defined as invasion of prostate cancer cells into the seminal vesicles. Presence or absence of ECE and SVI was recorded from the surgical pathology report.


Figure 9
View larger version (23K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 3 —Graph shows results of receiver operating characteristic analysis for detection of extracapsular extension with MRI alone and MRI with PACS cross-referencing. R1 = reviewer 1, R2 = reviewer 2, MRI = MRI alone, PACS = MRI with PACS cross-referencing, AUC = area under curve.

 
Statistical Analysis
Two of the authors were responsible for statistical analysis. Empirically estimated receiver operating characteristics (ROC) curves were used to evaluate the radiologists' accuracy in detecting SVI and ECE on MRI. For each ROC curve, the area under the curve (AUC) and the corresponding confidence intervals (CI) were estimated with nonparametric methods. Comparisons of the AUCs of a single reviewer were conducted with methods described by DeLong et al. [26]. To assess interobserver variability in determining the presence of ECE and SVI on MRI alone and MRI with PACS cross-referencing, we used a weighted kappa statistic with weights 1 - |i - j| / (5-1) where i, j = 1,..., 5 denotes the rating categories for the first and second reviewers [27]. Analyses were performed with Intercooled Stata 8.0 for Windows (Stata) and S-Plus for Windows 6.2.1 (Insightful).


Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Surgical Histopathologic Findings
At histopathologic examination, 68 (27%) of 255 patients had ECE and 13 (5%) had SVI. All 13 patients with SVI had ECE.

MRI Findings
Detection of ECE—ROC curves for reviewers 1 and 2 in detecting and localizing ECE are plotted in Figure 3. Reviewer 1 had an AUC of 0.66 (95% CI, 0.58-0.73) using MRI alone and an AUC of 0.87 (95% CI, 0.82-0.93) using MRI with PACS cross-referencing. Reviewer 1 performed significantly better using MRI with PACS cross-referencing than using MRI alone (p < 0.001). Reviewer 2 had an AUC of 0.69 (95% CI, 0.61-0.77) using MRI alone and an AUC of 0.86 (95% CI, 0.79-0.92) using MRI with PACS cross-referencing. Reviewer 2 also performed significantly better using MRI with PACS cross-referencing (p < 0.001). For detection of ECE with MRI alone, the weighted kappa was 0.56, and for detection of ECE with MRI and PACS cross-referencing, it was 0.76, indicating fair to good agreement between the two reviewers.

We dichotomized the 5-point scoring system and used 4 cut points to assess the sensitivity and specificity of MRI in the diagnosis of ECE (Table 1). When cut point 3 was chosen (so that values 1-3 indicated absence of ECE and values 4 and 5 indicated presence of ECE) for reviewer 1, sensitivity and specificity were 43% and 94% for MRI alone and 57% and 100% for MRI with PACS cross-referencing. At the same cut point for reviewer 2, sensitivity and specificity were 40% and 93% for MRI alone and 59% and 98% for MRI with PACS cross-referencing.


View this table:
[in this window]
[in a new window]

  		TABLE 1: Sensitivity and Specificity for Detection of Extracapsular Extension at 4 Cut Points of the 5-Point Scoring System

 

Detection of SVI—ROC curves for reviewers 1 and 2 in detecting and localizing SVI are plotted in Figure 4. Reviewer 1 had an AUC of 0.62 (95% CI, 0.48-0.76) using MRI alone and an AUC of 0.87 (95% CI, 0.77-0.96) using MRI with PACS cross-referencing. Reviewer 1 performed significantly better using PACS cross-referencing (p = 0.007). Reviewer 2 had an AUC of 0.73 (95% CI, 0.58-0.88) using MRI alone and an AUC of 0.90 (95% CI, 0.80-0.99) using MRI with PACS cross-referencing. Reviewer 2 also performed better using PACS cross-referencing, although for reviewer 2 the difference had only borderline significance (p = 0.056). In the detection of SVI, the weighted kappa statistic was 0.69 for both MRI alone and MRI with PACS cross-referencing, indicating good interobserver agreement.


Figure 10
View larger version (21K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 4 —Graph shows results of receiver operating characteristic analysis for detection of seminal vesicle invasion with MRI with PACS cross-referencing and MRI alone. R1 = reviewer 1; R2 = reviewer 2; MRI = MRI alone; PACS = MRI with PACS cross-referencing, AUC = area under curve.

 
We dichotomized the 5-point scoring system and used 4 cut points to assess the sensitivity and specificity of MRI in the diagnosis of SVI (Table 2). When cut point 3 was used (so that values 1-3 indicated absence of SVI and values 4 and 5 indicated presence of SVI) for reviewer 1, sensitivity and specificity were 23% and 83% for MRI alone and 46% and 93% for MRI with PACS cross-referencing. At the same cut point for reviewer 2, sensitivity and specificity were 31% and 91% for MRI alone and 54% and 95% for MRI with PACS cross-referencing.


View this table:
[in this window]
[in a new window]

  		TABLE 2: Sensitivity and Specificity for Seminal Vesicle Invasion at 4 Cut Points of the 5-Point Scoring System

 


Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
The Surveillance, Epidemiology, and End Results (SEER) program of the National Cancer Institute has shown a dramatic down-staging of prostate cancer at diagnosis over the last 2 decades. SEER data show that from 1995 to 2000, 90% of cases of prostate cancer were at a local or regional stage at diagnosis, compared with 72% from 1983 to 1987 [1, 3]. In our patient population, 27% of the patients had evidence of ECE and 5% had evidence of SVI at surgical histopathologic examination. These results are consistent with SEER data [2].

ECE and SVI are important prognostic factors for recurrence after radical prostatectomy [4, 28]. ECE is associated with greater risk of a positive surgical margin, which further decreases the chance of long-term cancer control [4, 7]. SVI is associated with an increased incidence of lymph node metastasis and a worse prognosis, even in the absence of lymph node involvement [29]. Accurate determination of the presence of ECE and SVI before treatment may alter treatment selection and planning [4, 7, 9].

MRI is generally considered the most accurate imaging method for local staging of prostate cancer [30-32]. Engelbrecht et al. [14] conducted a meta-analysis of 146 studies performed from January 1984 to May 2000 and described in 71 articles and five abstracts. The findings showed that MRI had an AUC of 0.60 ± 0.19 (SD) in detection of ECE and an AUC of 0.62 ± 0.13 in detection of SVI. The role of MRI in prostate cancer management has been changing with the development of techniques such as endorectal MRI, MR spectroscopic imaging, and PACS data storage [6, 10-12, 25, 33, 34]. With a PACS, medical information can be stored, recalled, displayed, manipulated, and printed digitally. Use of a PACS simplifies workflow, enhances productivity, and makes information accessible to multiple users simultaneously. Examination findings digitally stored in a secure computer system can be quickly transmitted to referring physicians, who in some cases can view the images in their offices via computer. Improvements in patient care include shorter hospital stays, decreased waiting times, faster diagnoses, and protection of personal medical information.

Our study showed that in the detection of ECE and SVI, radiologists performed substantially better using MRI with PACS cross-referencing than with MRI alone. PACS cross-referencing is particularly helpful in displaying the relationship between the seminal vesicles and the central zone of the prostate (Figs. 2A, 2B, 2C, and 2D). In our study, one reviewer benefited more from cross-referencing than did the other. This finding may relate to inherent differences in the way people see things.

One limitation of our study was that cross-referencing immediately followed the initial image review, so conclusions made at the initial review were incorporated into the final review. Thus there was almost no way that the reviewers could perform worse with cross-referencing than without it. Another limitation was that ECE and SVI were analyzed patient by patient but not according to specific location (anterior, lateral, posterior or apex, middle, base), so the ECE and SVI predicted were not conclusively and specifically located.

The initial sensitivities (43% and 40%) and AUC (0.66-0.69) of the reviewers for ECE in our study were relatively low for a major medical center with experienced reviewers [14]. The sensitivities (57% and 59%) and AUCs (0.87 and 0.86) after cross-referencing are more in line with expected findings. Both reviewers use the cross-referencing tool when routinely interpreting prostate MRI examinations. Therefore it is possible that when they were required not to use this tool at initial review, their accuracy suffered. It is also possible that the results were affected by anticipatory bias; that is, the radiologists may have expected better results with cross-referencing and held back in their initial interpretations without consciously intending to do so. In addition, the results may have been influenced by verification bias because the reviewers were aware that all patients had undergone surgical treatment, which was a decision influenced by initial MRI results.

The fact that all patients in this study underwent surgery implies that they had relatively low stages of local disease. Our study showed that cross-referencing on a PACS improved detection of relatively subtle ECE and SVI in these patients. In addition, the use of cross-referencing improved interobserver agreement in the detection of ECE. To support our findings and in light of the study limitations, we recommend further prospective multicenter confirmatory studies involving more reviewers with more varied experience.

In summary, our findings suggest that the PACS cross-referencing tool allows radiologists to more accurately interpret MR images of the prostate, significantly improving tumor staging of prostate cancer with MRI. Some reviewers benefit more than others from use of this tool.


Acknowledgments
 
We thank Ada Muellner for editing the manuscript.


References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Jemal A, Murray T, Ward E, et al. Cancer statistics, 2005. CA Cancer J Clin2005; 55:10 -30[Abstract/Free Full Text]
  2. American Cancer Society. Cancer facts & figures 2005. Atlanta, GA: American Cancer Society, 2005. Publication no. 5008.05
  3. Boring CC, Squires TS, Tong T. Cancer statistics, 1993. CA Cancer J Clin1993; 43:7 -26[Medline]
  4. Hull GW, Rabbani F, Abbas F, Wheeler TM, Kattan MW, Scardino PT. Cancer control with radical prostatectomy alone in 1,000 consecutive patients. J Urol 2002;167:528 -534[CrossRef][Medline]
  5. Catalona WJ, Ramos CG, Carvalhal GF. Contemporary results of anatomic radical prostatectomy. CA Cancer J Clin1999; 49:282 -296[Abstract]
  6. Hricak H, Wang L, Wei DC, et al. The role of preoperative endorectal magnetic resonance imaging in the decision regarding whether to preserve or resect neurovascular bundles during radical retropubic prostatectomy. Cancer2004; 100:2655 -2663[CrossRef][Medline]
  7. Ohori M, Kattan MW, Koh H, et al. Predicting the presence and side of extracapsular extension: a nomogram for staging prostate cancer. J Urol 2004;171:1844 -1849[CrossRef][Medline]
  8. Partin AW, Mangold LA, Lamm DM, Walsh PC, Epstein JI, Pearson JD. Contemporary update of prostate cancer staging nomograms (Partin tables) for the new millennium. Urology2001; 58:843 -848[CrossRef][Medline]
  9. Koh H, Kattan MW, Scardino PT, et al. A nomogram to predict seminal vesicle invasion by the extent and location of cancer in systematic biopsy results. J Urol2003; 170:1203 -1208[CrossRef][Medline]
  10. Wang L, Mullerad M, Chen HN, et al. Prostate cancer: incremental value of endorectal MR imaging findings for prediction of extracapsular extension. Radiology2004; 232:133 -139[Abstract/Free Full Text]
  11. Mullerad M, Hricak H, Wang L, Chen HN, Kattan MW, Scardino PT. Prostate cancer: detection of extracapsular extension by genitourinary and general body radiologists at MR imaging. Radiology2004; 232:140 -146[Abstract/Free Full Text]
  12. Cornud F, Flam T, Chauveinc L, et al. Extraprostatic spread of clinically localized prostate cancer: factors predictive of pT3 tumor and of positive endorectal MR imaging examination results. Radiology2002; 224:203 -210[Abstract/Free Full Text]
  13. Rajesh A, Coakley FV. MR imaging and MR spectroscopic imaging of prostate cancer. Magn Reson Imaging Clin N Am2004; 12:557 -579[CrossRef][Medline]
  14. Engelbrecht MR, Jager GJ, Laheij RJ, Verbeek AL, van Lier HJ, Barentsz JO. Local staging of prostate cancer using magnetic resonance imaging: a metaanalysis. Eur Radiol2002; 12:2294 -2302[Medline]
  15. Wallner K. MR imaging for prostate cancer staging: beauty or beast? Int J Radiat Oncol Biol Phys2002; 52:886 -887[CrossRef][Medline]
  16. Sonnad SS, Langlotz CP, Schwartz JS. Accuracy of MR imaging for staging prostate cancer: a metaanalysis to examine the effect of technologic change. Acad Radiol2001; 8:149 -157[CrossRef][Medline]
  17. May F, Treumann T, Dettmar P, Hartung R, Breul J. Limited value of endorectal magnetic resonance imaging and transrectal ultrasonography in the staging of clinically localized prostate cancer. BJU Int 2001;87:66 -69[CrossRef][Medline]
  18. Gray H, Lewis WH. Anatomy of the human body, 20th ed. Available at: www.bartleby.com. Accessed August 31, 2005
  19. Coakley FV, Hricak H. Radiologic anatomy of the prostate gland: a clinical approach. Radiol Clin North Am2000; 38:15 -30[CrossRef][Medline]
  20. Gonzalez Ballester MA, Zisserman AP, Brady M. Estimation of the partial volume effect in MRI. Med Image Anal2002; 6:389 -405[CrossRef][Medline]
  21. Maeda M, Suzuki E, Yoshiya K, et al. Partial volume effect in MRI: a phantom study [in Japanese]. Nippon Igaku Hoshasen Gakkai Zasshi 1989;49:1404 -1410[Medline]
  22. Miyamoto K, Abe S, Kawakami Y. Picture archiving and communication system in Hokkaido University Hospital: advantage and disadvantage of HU-PACS chest roentgenogram images in the outpatient clinic. J Digit Imaging 1991;4:28 -31[Medline]
  23. De Backer AI, Mortele KJ, De Keulenaer BL. Picture archiving and communication system. Part 1. Filmless radiology and distance radiology. JBR-BTR 2004;87:234 -241
  24. Hertzberg BS, Kliewer MA, Paulson EK, et al. PACS in sonography: accuracy of interpretation using film compared with monitor display—picture archiving and communication systems. AJR 1999;173:1175 -1179[Abstract/Free Full Text]
  25. Claus FG, Hricak H, Hattery RR. Pretreatment evaluation of prostate cancer: role of MR imaging and 1H MR spectroscopy. RadioGraphics2004; 24[suppl 1]:S167 -S180[Abstract/Free Full Text]
  26. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics1988; 44:837 -845[CrossRef][Medline]
  27. Cicchetti DV, Fleiss JL. A new procedure for assessing reliability if scoring EEG sleep recordings. Am J EEG Technol1977; 11:101 -109
  28. D'Amico AV, Renshaw AA, Sussman B, Chen MH. Pretreatment PSA velocity and risk of death from prostate cancer following external beam radiation therapy. JAMA2005; 294:440 -447[Abstract/Free Full Text]
  29. Epstein JI, Partin AW, Potter SR, Walsh PC. Adenocarcinoma of the prostate invading the seminal vesicle: prognostic stratification based on pathologic parameters. Urology2000; 56:283 -288[CrossRef][Medline]
  30. Ikonen S, Karkkainen P, Kivisaari L, et al. Endorectal magnetic resonance imaging of prostatic cancer: comparison between fat-suppressed T2-weighted fast spin echo and three-dimensional dual-echo, steady-state sequences. Eur Radiol2001; 11:236 -241[CrossRef][Medline]
  31. Rorvik J, Halvorsen OJ, Albrektsen G, Ersland L, Daehlin L, Haukaas S. MRI with an endorectal coil for staging of clinically localised prostate cancer prior to radical prostatectomy. Eur Radiol1999; 9:29 -34[CrossRef][Medline]
  32. Tempany CM. MR staging of prostate cancer: how we can improve our accuracy with decisions aids and optimal techniques. Magn Reson Imaging Clin N Am 1996;4:519 -532[Medline]
  33. Futterer JJ, Scheenen TW, Huisman HJ, et al. Initial experience of 3 tesla endorectal coil magnetic resonance imaging and 1H-spectroscopic imaging of the prostate. Invest Radiol2004; 39:671 -680[CrossRef][Medline]
  34. Sosna J, Pedrosa I, Dewolf WC, Mahallati H, Lenkinski RE, Rofsky NM. MR imaging of the prostate at 3 Tesla: comparison of an external phased-array coil to imaging with an endorectal coil at 1.5 Tesla. Acad Radiol2004; 11:857 -862[CrossRef][Medline]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 Am. J. Roentgenol.Home page
R. J. Stanley
The Challenge of Managing Prostate Cancer
Am. J. Roentgenol., January 1, 2007; 188(1): 1 - 1.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wang, L.
Right arrow Articles by Hricak, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wang, L.
Right arrow Articles by Hricak, H.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS