Published online before print January 14, 2008, 10.1148/radiol.2463070221
(Radiology 2008;246:804-811.)
© RSNA, 2008
USPIO-enhanced MR Imaging for Nodal Staging in Patients with Primary Rectal Cancer: Predictive Criteria1
Max J. Lahaye, MD,
Sanne M. E. Engelen, MD,
Alfons G. H. Kessels, MD, PhD,
Adriaan P. de Bruïne, MD, PhD,
Maarten F. von Meyenfeldt, MD, PhD,
Jos M. A. van Engelshoven, MD, PhD,
Cornelis J. H. van de Velde, MD, PhD,
Geerard L. Beets, MD, PhD, and
Regina G. H. Beets-Tan, MD, PhD
1 From the Departments of Radiology (M.J.L., S.M.E.E., J.M.A.v.E., R.G.H.B.), Surgery (M.J.L., S.M.E.E., M.F.v.M., G.L.B.), Epidemiology (A.G.H.K.), and Pathology (A.P.d.B.), University Hospital Maastricht, P. Debyelaan 25, 6229 HX Maastricht, PO Box 5800, 6202 AZ Maastricht, the Netherlands; and Department of Surgery, Leiden University Medical Center, Leiden, the Netherlands (C.J.H.v.d.V.). Received February 1, 2007; revision requested April 3; revision received May 14; accepted May 30; final version accepted September 11. Supported by the Dutch Cancer Society.
Address correspondence to M.J.L. (e-mail: MLAHAYE{at}rdia.azm.nl).
 |
ABSTRACT
|
|---|
Purpose: To prospectively determine diagnostic performance of predictive criteria for nodal staging with ultrasmall superparamagnetic iron oxide (USPIO)–enhanced magnetic resonance (MR) imaging in primary rectal cancer patients, with histopathologic findings as reference standard.
Materials and Methods: Institutional review board approval and informed consent were obtained. Twenty-eight rectal cancer patients (18 men, 10 women; mean age, 68 years) underwent USPIO-enhanced MR. Two observers with different experience evaluated each node on three-dimensional T2*-weighted images for border irregularity, short- and long-axis diameter, and estimated percentage (<30%, 30%–50%, or >50%) of white region within the node. Ratio of measured surface area of white region within the node to measured surface area of total node (ratioA) was calculated. Signal intensity (SI) of gluteus muscle (SIGM), total node (SITN), and white (SIWR) and dark (SIDR) regions within the node were used to calculate SITN/SIGM and SIWR/SIDR ratios. Lesion-by-lesion, receiver operating characteristic curve, and interobserver agreement analyses were performed. The most accurate and practical criterion was evaluated by observer 3.
Results: In 28 patients, 236 lymph nodes were examined. Area under the receiver operating characteristic curve (AUC) of estimated percentage of white region and ratioA were 0.96 and 0.99 (observer 1) and 0.95 and 0.97 (observer 2), respectively. AUC of estimated percentage criterion for observer 3 was 0.96. AUC of border, short- and long-axis diameter, and SITN/SIGM and SIWR/SIDR ratios were 0.65, 0.75, 0.79, 0.85, and 0.75 (observer 1) and 0.58, 0.75, 0.79, 0.89, and 0.79 (observer 2), respectively. Interobserver agreement (
value) for estimated white region between observers 1 and 2, 1 and 3, and 2 and 3 were 0.77, 0.79, and 0.84, respectively. For observers 1 and 2, value for border was 0.28. For observers 1 and 2, intraclass correlation coefficient for short- and long-axis diameters, ratioA, and SITN/SIGM and SIWR/SIDR ratios were 0.91, 0.96, 0.91, 0.72, and 0.92, respectively.
Conclusion: Estimated percentage of white region and measured ratioA are the most accurate criteria for predicting malignant nodes with USPIO-enhanced MR imaging; the first criterion is the most practical.
© RSNA, 2008
|
INTRODUCTION
|
|---|
Rectal cancer is a common malignancy with a highly variable local recurrence rate of 3%–32% (1). In the past, besides total mesorectal excision, neoadjuvant rather than adjuvant radiation therapy and chemotherapy were advocated to reduce the local recurrence rate (2,3). Furthermore, to prevent over- and undertreatment, a tailor-made neoadjuvant approach for every rectal cancer patient that is based on preoperative prediction of the risk for local recurrence seems to be favorable. Important risk factors for local recurrences are tumor stage, circumferential resection margin, nodal status, and height of the tumor. Preoperative imaging is essential to identify these risk factors.
The nodal status is a well-established important prognostic risk factor. The higher the number of nodes positive for cancer, the higher the risk for a local recurrence (4,5). However, prediction of the nodal status remains a problem for the radiologist. The results of a meta-analysis indicate that, for identification of nodal disease, endoluminal ultrasonography (US), magnetic resonance (MR) imaging, and computed tomography (CT) lack sufficient accuracy for clinical decision making (6).
Ultrasmall superparamagnetic iron oxide (USPIO)–enhanced MR imaging has been reported to be promising for differentiation of benign from malignant nodes. Results of a recent meta-analysis indicated that USPIO-enhanced MR imaging is superior to all other modalities in the detection of lymph node metastases for various tumors (7). Despite many promising results concerning USPIO-enhanced MR imaging, there is still insufficient knowledge of which specific USPIO-enhanced MR imaging criteria can be used for characterizing lymph nodes as benign or malignant. Thus, the purpose of our study was to prospectively determine the diagnostic performance of various predictive criteria for nodal staging with USPIO-enhanced MR imaging in primary rectal cancer patients, with histopathologic findings as the reference standard.
|
MATERIALS AND METHODS
|
|---|
The study was financed by the Dutch Cancer Society. Guerbet Laboratories (Roissy, France) provided vials of the contrast agent used in this study (Sinerem). The authors had control of the data and information submitted for publication.
Patients
Institutional review board approval and informed consent were obtained. During the study period of June 2004 to July 2006, a total of 87 consecutive patients received a diagnosis of biopsy-proved primary rectal cancer at University Hospital Maastricht, Maastricht, the Netherlands. All rectal tumors were defined by an inferior tumor margin of less than 15 cm from the anal verge, as described by using endoscopy. Fourteen patients with nonresectable disease or contraindications for MR imaging were excluded from the study. Forty-five patients with locally advanced tumors (involved circumferential resection margin or N2 status) underwent neoadjuvant chemotherapy and radiation therapy and were excluded because, after chemotherapy and radiation therapy, histopathologic findings cannot be considered representative of the primary nodal status. Twenty-eight patients were eligible for this prospective study (18 men, 10 women; mean age, 68 years; range, 54–89 years). These patients underwent USPIO-enhanced MR imaging before their standard treatment. The standard treatment in our country for nonlocally advanced rectal cancer is a short course of radiation therapy (5 Gy of radiation therapy on 5 consecutive days), immediately followed by total mesorectal excision (8). The short course of radiation therapy is known not to lead to downstaging of the nodal status in rectal cancer patients (9).
Ultrasmall Superparamagnetic Iron Oxide
The USPIO MR contrast agent used in this study consists of low–molecular-weight iron oxide coated with dextran, is supplied as a powder in a glass vial containing 210 mg, and must be reconstituted by using 10 mL of normal saline. A dose of 0.13 mL per kilogram of body weight (2.6 mg of iron per kilogram) of the reconstituted solution was diluted in 100 mL of normal saline. The contrast agent was administered intravenously in the preparation room of the MR unit within a period of approximately 45 minutes by means of a slow-drip infusion with a microfilter. Administration was closely monitored for any adverse effects and was completed 24–36 hours before contrast material–enhanced MR imaging was planned. No unenhanced MR imaging was performed. No adverse events relating to the USPIO infusion were observed in this study.
In the literature, two main mechanisms for transportation of the nanoparticles to the lymph nodes are described. The major pathway is direct transcapillary passage of the nanoparticles from blood vessels into the medullary sinuses of the lymph node. The other pathway is nonselective endothelial transcytosis across permeable capillaries into the interstitium, from where nanoparticles are transported through the lymphatic system and accumulate in macrophages within the lymph node (10,11). The nanoparticles cause a decrease in signal intensity (SI) within the node owing to susceptibility artifacts on three-dimensional (3D) T2*-weighted MR images. Variation of SI within a node can be explained on the basis of the concentration of macrophages (nanoparticles) in a particular region in the node. A region with a normal concentration of macrophages (nanoparticles) will produce an area with SI decrease. This means that enlarged inflamed nodes will also show a significant decrease in SI. The involved part of the lymph node will show no SI decrease caused by the replacement of macrophages by tumor cells, creating a region of increased SI within the node (white region).
MR Imaging
MR imaging was performed with a 1.5-T system (Intera; Philips Medical Systems, Best, the Netherlands) by using a cardiac phased-array coil with the patient in the feet-first supine position, with a gradient strength of 23.0 mT/m, a rise time of 0.2 msec, and a slew rate of 105 T/m/sec. Sequences used were sagittal and transverse two-dimensional T2-weighted fast spin echo. Transverse 3D T2*-weighted MR images were obtained with repetition time msec/echo time msec, 23/18.41; flip angle, 20°; matrix, 512 x 512; and section thickness, 1.25 mm. Transverse 3D T1-weighted MR images were obtained with 10/4.60; flip angle, 15°; matrix, 384 x 512; and section thickness, 1 mm. The latter sequence was used to locate lymph nodes and to differentiate them from blood vessels. All transverse images were angled perpendicular to the rectal tumor. Patients did not receive bowel or other preparation. The total image time was 32–40 minutes.
Image Evaluation
Two observers (R.G.H.B. [observer 1] and M.J.L. [observer 2], with experience in reading approximately 3000 and 300 pelvic MR images, respectively), who were blinded to each other's results and histopathologic results, prospectively evaluated each node on transverse 3D T2*-weighted MR images for the following items: border (sharp, indistinct, or disrupted) and short- and long-axis diameter. Furthermore, the percentage of white region within the node was estimated and categorized as less than 30%, 30%–50%, or more than 50% on USPIO-enhanced 3D T2*-weighted MR images (Fig 1).
View larger version (7K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1: Schematic of a lymph node on USPIO-enhanced 3D T2*-weighted MR images. Left: White region of less than 30%. Middle: White region of 30%–50%. Right: White region of more than 50%.
|
|
Besides this visual assessment of the percentage of white region within the node, the proportion of white region within the node also was quantitatively measured. The surface area of the white region within the node and the surface area of the total node were determined on a transverse 3D T2*-weighted MR image, where the node was the largest, by using the measurement tools of the workstation. By dividing the surface area of the white region by the surface area of the total node, a quantitatively measured ratio of white region was calculated for both observers. This is further referred to as the ratio of the measured surface area of the white region within the node to the measured surface area of the total node (ratioA), or the MSAWR/MSATN ratio, where MSAWR is the measured surface area of the white region within the node and MSATN is the measured surface area of the total node. RatioA was not categorized.
By placing regions of interest completely over the white and dark regions within the node, the SI values were measured. Regions of interest of the same size were used to measure the SI of the total node (SITN) and SI of the gluteus muscle (SIGM). These SI values were used to calculate the following two ratios for both observers: ratio of SI of the white region within the node (SIWR) to SI of the dark region within the node (SIDR), or SIWR/SIDR, and SITN/SIGM.
The observer with the highest level of experience, observer 1, also prospectively recorded the localization of each visible lymph node depicted by using 3D T2*-weighted MR images on an anatomic map (Fig 2). This anatomic map was used as a template to compare with the nodes detected by the other observers (as indicated later) and the nodes found at histopathologic evaluation.
View larger version (49K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2: Each node found at histopathologic evaluation was matched with the node on the MR image. Left: First, a node (arrow) was predicted on MR image. Middle: This node (arrow) was recorded on an anatomic map at corresponding height (I–V). Right: Node (arrow) was harvested from histologic specimen and matched with node depicted on MR image, with anatomic map used as template.
|
|
A third observer was introduced to evaluate the most accurate and practical predictive criterion for nodal status of observers 1 and 2. Observer 3 (S.M.E.E.) had reading experience similar to that of observer 2 (approximately 300 pelvic MR image readings).
Histopathologic Evaluation
The standard surgical procedure was total mesorectal excision, as described by Heald (12). The rectum with the complete surrounding mesorectal fat and mesorectal lymph nodes was removed by sharp dissection along the mesorectal fascia. Histopathologic examination was standardized: The specimen was fixed in formalin for 24–48 hours, and the circumferential resection plane was inked. The specimen was then sectioned every 5 mm transversely and, thus, also perpendicularly to the mesorectum. A careful search for lymph nodes was made in each histologic tissue slice by a dedicated pathologist. Each lymph node found was matched to the corresponding node visible on the MR images by using the anatomic map as a template. This nodal harvesting procedure was performed by a pathologist (A.P.d.B., with 17 years of experience) together and side by side with observer 1 to ensure an accurate lesion-by-lesion analysis. Accurate nodal matching was only possible because of the detailed knowledge of the MR images by using specific characteristics, such as the size of the lymph node and the position of the lymph node in relation to the rectal wall, small blood vessels, mesorectal fascia, tumor, and other lymph nodes. In cases in which nodal matching was not possible, the cases were recorded as such. Each harvested lymph node was placed in a marked individual tray and was processed according to standard methods and stained with hematoxylin-eosin. Then, the same pathologist, who was blinded to predictions on the basis of the USPIO-enhanced MR imaging findings, reported the status of each lymph node: benign or malignant. In this manner, accurate lesion-by-lesion analysis for mesorectal lymph nodes could be performed.
Statistical Analysis
On the basis of the results of a pilot study at our center, we expected a sensitivity of 99%. Thirty lymph nodes positive for cancer are needed to reject, with a power of 75%, that the lower limit of the 95% confidence interval of sensitivity of USPIO-enhanced MR imaging is 90% or less. The number of patients needed to include 30 lymph nodes positive for cancer can be calculated: Ten percent of the nodes found will be malignant, and an average of 12 nodes in each patient will be found at histopathologic examination; this means that we need 300 nodes at histopathologic examination and, thus, 25 patients.
To determine the interobserver agreement, weighted values (13,14) were calculated for the ordinal variables, and intraclass correlation coefficients were calculated for interval-scaled variables (15), all with their 95% confidence intervals. Agreement was analyzed with the value as follows: A value of less than 0 indicated poor agreement; 0–0.20, slight agreement; 0.21–0.40, fair agreement; 0.41–0.60, moderate agreement; 0.61–0.80, substantial agreement; and 0.81–1.00, almost perfect agreement (16). The intraclass correlation coefficient ranges from zero to one, with a value of zero indicating a completely unreliable measurement and a value of one indicating a 100% reproducible measurement.
Diagnostic performance of the different evaluation methods was analyzed with the logistic regression model, and generalized estimating equations were used to estimate marginal effects. With the results, receiver operating characteristic curves were constructed, and area under the receiver operating characteristic curve (AUC) and 95% confidence intervals were calculated by using software (Stata, release 9; StataCorp, College Station, Tex).
|
RESULTS
|
|---|
Histopathologic Findings
In 28 patients, MR imaging aided in finding 362 lymph nodes in the mesorectum. After surgical resection, 333 lymph nodes were found at histopathologic evaluation; 27 lymph nodes were malignant and 306 were benign. Of the 333 (sum of 27 and 306) nodes found at histopathologic evaluation, 236 nodes could be matched exactly with nodes found on MR images, whereas 97 nodes could not. These 97 nodes, all benign, were excluded from the analysis. This left 236 lymph nodes for the lesion-by-lesion analysis (Fig 3). Only 176 nodes were eligible for SI measurements, because 60 nodes were too small to accurately measure the SI of the node. The mean short- and long-axis diameters of the total lymph node were 3.6 mm (range, 2–9 mm) and 4.7 mm (range, 2–13 mm), respectively.
Diagnostic Performance
The AUC of border, short- and long-axis diameters, and SITN/SIGM and SIWR/SIDR ratios for observer 1 were 0.65, 0.75, 0.79, 0.85, and 0.75, respectively, and for observer 2 were 0.58, 0.75, 0.79, 0.89, and 0.79, respectively (Table 1, Fig 4).
View this table:
[in this window]
[in a new window]
|
Table 1. AUC and 95% Confidence Intervals for Observers 1 and 2 for All Criteria Used to Distinguish between Malignant and Benign Nodes
|
|
View larger version (28K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4: Receiver operating characteristic curves and AUCs for observers 1 and 2 for detection of malignant lymph nodes by using short- and long-axis diameter and border of lymph node.
|
|
The highest AUCs for the prediction of the nodal status were found for the estimated percentage and measured ratioA of the white region within the node (0.96 and 0.99, respectively, for observer 1 and 0.95 and 0.97, respectively, for observer 2). When the white region within the node is estimated as more than 30%, the sensitivity and specificity for detection of malignant nodes are 93% and 96%, respectively (Figs 5, 6).
View larger version (24K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5: Receiver operating characteristic curves and AUCs (observers 1–3) for detection of malignant lymph nodes by using estimated percentage of white region within the node and ratioA (Ratio) (observers 1 and 2).
|
|
View larger version (148K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6: A–D, Benign mesorectal node diagnosed on USPIO-enhanced MR images. A, On T2-weighted fast spin-echo image, node (open rectangle) was smaller than 5 mm and it was difficult to differentiate whether node was malignant or benign. B, On USPIO-enhanced T2*-weighted MR image, node caused black blooming artifact. C, Less than 30% white region within the node, indicating a benign node. D, Histologic evaluation findings confirmed node to be benign. E–H, Small malignant mesorectal node detected with USPIO-enhanced MR imaging. E, On T2-weighted fast spin-echo MR image, node (open rectangle) was smaller than 5 mm and it was difficult to differentiate whether node was malignant or benign. F, On USPIO-enhanced T2*-weighted MR image, incomplete blackening of the node was observed. G, Estimated 30%–50% white region within the node, suggesting a malignant node. H, Histologic evaluation findings confirmed partially involved metastatic node. I–L, Ultrasmall malignant mesorectal node detected with USPIO-enhanced MR imaging. I, On T2-weighted fast spin-echo MR image, node (open rectangle) was smaller than 3 mm and it was difficult to differentiate whether the node was malignant or benign. J, On USPIO-enhanced T2*-weighted MR image, remaining white area was observed, indicating no iron uptake. K, More than 50% white area within the node, suggesting a malignant node. L, Histologic evaluation findings confirmed a massively involved tumoral node.
|
|
Of the two criteria that were the most accurate in the prediction of the nodal status with USPIO-enhanced MR imaging, the estimated percentage of white region within the node was the most practical to use in clinical practice. Therefore, this criterion was chosen to be evaluated by observer 3. The AUC of observer 3 for this criterion was 0.96 (95% confidence interval: 0.91, 1.00).
Intraclass Correlation Coefficient
The intraclass correlation coefficient for observers 1 and 2 for the short- and long-axis diameters, ratioA, and SITN/SIGM and SIWR/SIDR ratios were 0.91, 0.96, 0.91, 0.72, and 0.92, respectively (Table 2). Generally, the interobserver agreement was good to excellent. There was more agreement for ratioA and short- and long-axis diameters than there was for the SITN/SIGM ratio.
View this table:
[in this window]
[in a new window]
|
Table 2. Intraclass Correlation Coefficients and 95% Confidence Intervals for All Interval-scaled Criteria for Observers 1 and 2
|
|
Interobserver Agreement
The interobserver agreement ( value) for the estimated white region between observers 1 and 2 was good (0.77), whereas that for the border was only fair (0.28). The values for the estimated white region for observers 1 and 3 and observers 2 and 3 were good (0.79) and excellent (0.84), respectively.
|
DISCUSSION
|
|---|
Our results indicate that, for observers 1 and 2, both the estimated and the measured ratio of the white region within the node were very accurate predictors for malignant nodes in rectal cancer. The results of observer 3 confirmed that the estimated percentage of the white region within the node could serve as a reliable and practical criterion for the prediction of malignant nodes on USPIO-enhanced MR images.
Size criteria, such as long- and short-axis diameters of the node, and border morphology were only moderately accurate for the distinction between malignant and benign lymph nodes.
An estimated area of white region within the node that was larger than 30% was highly predictive for an involved node, with a sensitivity of 93% and a specificity of 96%. The measured ratioA was also an accurate predictor for malignant lymph nodes. The larger the area of the white region, the more likely the node is malignant. The white region in the lymph node is caused by no or very little uptake of USPIO in that malignant part of the lymph node. Benign conditions such as focal nodal fibrosis, granulomatous disease, or a fatty hilum also can be depicted as a white region because of the lack of macrophages, thus mimicking malignant nodes. These white regions, however, are usually 30% or less of the total node area. Another way to differentiate between a white fatty hilum and a white tumoral region is to compare the 3D T2*-weighted images with the 3D T1-weighted images: A fatty hilum is depicted as a white region within the lymph node, whereas a tumor is not. In the same manner, the dark region can hide small micrometastases. The clinical implication of micrometastases, however, is debatable. Although there is a small overlap in MR features of benign and malignant nodes, a general rule of thumb is that, when the white region is less than 30% of the total node area, the nodes are most often benign.
To our knowledge, no literature has been published concerning the estimated or measured ratio of the white region within the node. A range of patterns of contrast enhancement has been described to discriminate between malignant and nonmalignant lymph nodes by using USPIO-enhanced MR images (17). The use of patterns to predict malignant nodes has a major drawback. The radiologist must choose from a variety of patterns and assess which patterns would fit best to a malignant lymph node and which would fit best to a benign node. This involves a certain level of subjectivity and constitutes a potential source of erroneous interpretation. Our study findings show that a simple estimation or measurement of the white region within the node is reliable for the identification of malignant nodes. The interobserver agreement between an experienced reader and less experienced readers (observer 1 vs observer 2 or 3) for the estimated percentage of white region were good, indicating a high consistency in interpreting the images independent of the experience of the readers. The value of 0.84 for observer 2 versus 3 even shows that two less experienced readers can have an excellent agreement between each other. These findings suggest that USPIO-enhanced MR imaging could become a promising tool for prediction of malignancy of nodes not only for experienced but also for less experienced readers.
The SITN/SIGM ratio measurements indicate that the higher the SITN, the higher the risk of a malignant node. These results confirm findings in earlier reports of SI values on USPIO-enhanced MR images. Harisinghani et al (18) demonstrated that a distinction between malignant and benign nodes could be made on the basis of SITN. In our study, the SIWR/SIDR ratio measurements were disappointing. The moderate AUC can be explained by the difficulties of measuring a white and a dark region within a very tiny node. The white and dark regions within the node are often too small or too irregular to accurately position a region of interest. In this study, 25% (60 of 236) of all nodes could not be measured, making this criterion less valuable in clinical practice.
Size criteria were insufficient to consistently distinguish between malignant and benign lymph nodes in patients with rectal cancer. The disappointing results concerning the size criteria were caused by the high rate of relatively small involved lymph nodes in rectal cancer. Brown et al (19) showed that even mesorectal lymph nodes smaller than 5 mm in diameter still had a prevalence of lymph nodes that were positive for cancer of 15%. As early as 1989, Dworak (20) showed that 32% of the patients with involved nodes had only small (
5- mm-diameter) involved nodes. This partly explains why endoluminal US, CT, and unenhanced MR studies for prediction of malignant lymph nodes in rectal cancer patients by using size criteria have shown disappointing results (6). Additional MR criteria such as border and SI have, therefore, been studied. Kim et al (21) and Brown et al (19) found that the border and SI characteristics of lymph nodes on T2-weighted fast spin-echo MR images were useful for identification of malignant nodes in rectal cancer. The aspect of the border and the homogeneity of the SI, however, are generally easier to evaluate in larger nodes (>5 mm in diameter). In our study, however, we mainly included patients with nonlocally advanced tumors with predominantly smaller nodes. In our experience, USPIO-enhanced MR imaging and its characteristics are especially of additional benefit in the evaluation of these small nodes (<5 mm in diameter), where border and SI criteria tend to be less helpful.
A limitation of our study is the relatively limited number of nodes positive for malignancy. This is due to the study's patient selection of only "nonlocally advanced" rectal cancers. These patients do not have that many malignant nodes and are stratified for a short preoperative radiation therapy course. An accurate lesion-by-lesion analysis can be performed in these patients, because, in contrast to neoadjuvant chemotherapy and radiation therapy, short-term preoperative radiation therapy does not lead to downstaging of malignant nodes (9).
In conclusion, the estimated percentage of white region within the node and ratioA are the most accurate predictive criteria for differentiation between benign and malignant nodes with USPIO-enhanced MR imaging. In the busy daily practice of a radiologist, there is little room for time-consuming measurements of numerous nodes. Therefore, the estimated percentage of white region within the node would be most practical to use. Our study results indicate that this was an accurate criterion not only for an experienced reader but also for less experienced readers.
|
ADVANCE IN KNOWLEDGE
|
|---|
- The most accurate and practical predictive criterion for malignant lymph nodes on USPIO-enhanced MR images in patients with primary rectal cancer is the percentage of white region within the node.
|
IMPLICATION FOR PATIENT CARE
|
|---|
- Our results suggest that USPIO-enhanced MR imaging could become a promising tool for nodal malignancy prediction not only for experienced but also for less experienced readers.
|
FOOTNOTES
|
|---|
Abbreviations: AUC = area under the receiver operating characteristic curve • ratioA = ratio of the measured surface area of the white region within the node to the measured surface area of the total node • SI = signal intensity • SIDR = SI of the dark region within the node • SIGM = SI of the gluteus muscle • SITN = SI of the total node • SIWR = SI of the white region within the node • 3D = three-dimensional • USPIO = ultrasmall superparamagnetic iron oxide
See Materials and Methods for pertinent disclosures.
Author contributions: Guarantor of integrity of entire study, R.G.H.B.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, all authors; clinical studies, M.J.L., S.M.E.E., A.P.d.B., M.F.v.M., J.M.A.v.E., C.J.H.v.d.V., G.L.B., R.G.H.B.; statistical analysis, M.J.L., A.G.H.K.; and manuscript editing, all authors
|
References
|
|---|
- Sagar PM, Pemberton JH. Surgical management of locally recurrent rectal cancer. Br J Surg 1996;83:293–304.[Medline]
- Pahlman L, Glimelius B. Pre- or postoperative radiotherapy in rectal and rectosigmoid carcinoma: report from a randomized multicenter trial. Ann Surg 1990;211:187–195.[Medline]
- Sauer R, Becker H, Hohenberger W, et al. Preoperative versus postoperative chemoradiotherapy for rectal cancer. N Engl J Med 2004;351:1731–1740.[Abstract/Free Full Text]
- Moran MR, James EC, Rothenberger DA, Goldberg SM. Prognostic value of positive lymph nodes in rectal cancer. Dis Colon Rectum 1992;35:579–581.[CrossRef][Medline]
- Park YJ, Park KJ, Park JG, Lee KU, Choe KJ, Kim JP. Prognostic factors in 2230 Korean colorectal cancer patients: analysis of consecutively operated cases. World J Surg 1999;23:721–726.[CrossRef][Medline]
- Lahaye MJ, Engelen SM, Nelemans PJ, et al. Imaging for predicting the risk factors–the circumferential resection margin and nodal disease–of local recurrence in rectal cancer: a meta-analysis. Semin Ultrasound CT MR 2005;26:259–268.[CrossRef][Medline]
- Will O, Purkayastha S, Chan C, et al. Diagnostic precision of nanoparticle-enhanced MRI for lymph-node metastases: a meta-analysis. Lancet Oncol 2006;7:52–60.[CrossRef][Medline]
- Kapiteijn E, Marijnen CA, Nagtegaal ID, et al. Preoperative radiotherapy combined with total mesorectal excision for resectable rectal cancer. N Engl J Med 2001;345:638–646.[Abstract/Free Full Text]
- Marijnen CA, Nagtegaal ID, Klein Kranenbarg E, et al. No downstaging after short-term preoperative radiotherapy in rectal cancer patients. J Clin Oncol 2001;19:1976–1984.[Abstract/Free Full Text]
- Weissleder R, Elizondo G, Wittenberg J, Rabito CA, Bengele HH, Josephson L. Ultrasmall superparamagnetic iron oxide: characterization of a new class of contrast agents for MR imaging. Radiology 1990;175:489–493.[Abstract/Free Full Text]
- Frija G, Clement O, Le Guen O, Cuenod CA, Siauve N, Benderbous S. Experimental investigation of the delivery pathway of ultrasmall superparamagnetic iron oxide to lymph nodes. Acad Radiol 1996;3(suppl 2):S299–S300.[CrossRef][Medline]
- Heald RJ. A new approach to rectal cancer. Br J Hosp Med 1979;22:277–281.[Medline]
- Cohen J. A coefficient of agreement for nominal scales. Educ Psychol Meas 1960;20:37–46.[CrossRef]
- Cohen J. Weighted kappa: nominal scale agreement with provision for scaled disagreement or partial credit. Psychol Bull 1968;70:213–230.[CrossRef]
- Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull 1979;86:420–428.[CrossRef][Medline]
- Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159–174.[CrossRef][Medline]
- Koh DM, Brown G, Temple L, et al. Rectal cancer: mesorectal lymph nodes at MR imaging with USPIO versus histopathologic findings–initial observations. Radiology 2004;231:91–99.[Abstract/Free Full Text]
- Harisinghani MG, Saini S, Slater GJ, Schnall MD, Rifkin MD. MR imaging of pelvic lymph nodes in primary pelvic carcinoma with ultrasmall superparamagnetic iron oxide (Combidex): preliminary observations. J Magn Reson Imaging 1997;7:161–163.[Medline]
- Brown G, Richards CJ, Bourne MW, et al. Morphologic predictors of lymph node status in rectal cancer with use of high-spatial-resolution MR imaging with histopathologic comparison. Radiology 2003;227:371–377.[Abstract/Free Full Text]
- Dworak O. Number and size of lymph nodes and node metastases in rectal carcinomas. Surg Endosc 1989;3:96–99.[CrossRef][Medline]
- Kim JH, Beets GL, Kim MJ, Kessels AG, Beets-Tan RG. High-resolution MR imaging for nodal staging in rectal cancer: are there any criteria in addition to the size? Eur J Radiol 2004;52:78–83.[CrossRef][Medline]
This article has been cited by other articles:
|
|
|
|
 
M. J. Lahaye, G. L. Beets, S. M. E. Engelen, A. G. H. Kessels, A. P. de Bruine, H. W. S. Kwee, J. M. A. van Engelshoven, C. J. H. van de Velde, and R. G. H. Beets-Tan
Locally Advanced Rectal Cancer: MR Imaging for Restaging after Neoadjuvant Radiation Therapy with Concomitant Chemotherapy Part II. What Are the Criteria to Predict Involved Lymph Nodes?
Radiology,
July 1, 2009;
252(1):
81 - 91.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
B. Barbaro, C. Fiorucci, C. Tebala, V. Valentini, M. A. Gambacorta, F. M. Vecchio, G. Rizzo, C. Coco, A. Crucitti, C. Ratto, et al.
Locally Advanced Rectal Cancer: MR Imaging in Prediction of Response after Preoperative Chemotherapy and Radiation Therapy
Radiology,
March 1, 2009;
250(3):
730 - 739.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
