| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Musculoskeletal Imaging |
1 From the Service de Radiologie B, Service de Chirurgie Orthopédique, Hôpital Cochin, 27 rue du Faubourg Saint Jacques, 75679 Paris Cedex 14, France. From the 2000 RSNA Annual Meeting. Received October 7, 2004; revision requested December 15; revision received February 22, 2005; accepted March 15; final version accepted April 15. Address correspondence to A.F. (e-mail: antoine.feydy{at}cch.aphp.fr).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Materials and Methods: This study was approved by the regional ethics committee, and all patients gave informed consent. Preoperative MR angiograms and MR images of 31 sites in 30 patients with bone or soft-tissue sarcomas (n = 21) or other tumors (n = 9) were assessed for features of vascular invasion. All images were prospectively evaluated by two musculoskeletal radiologists working in consensus. MR images were evaluated for the presence of a fat plane between the vessels and the tumor and partial or total encasement of vessels. MR angiograms were evaluated for the presence of vascular displacement, stenosis, or occlusion. MR imaging and MR angiographic features of vascular invasion were graded. Imaging findings were correlated with surgical findings and classified as negative if there was no vascular invasion and as positive if there was vascular invasion.
Results: Among the 31 cases, 20 were classified as negative and 11 were classified as positive at surgery. All but three cases with a gap between the tumor and the vessels on MR images were classified as free and without adhesions at surgery. All cases with arterial stenoses at MR angiography had tumoral adhesion or tumoral encasement at surgery. MR imaging had a sensitivity of 64%, a specificity of 95%, a positive predictive value of 88% a negative predictive value of 83%, and an accuracy of 84% in the detection of vascular invasion on the basis of findings of partial or total encasement. MR angiography had a sensitivity of 82%, a specificity of 85%, a positive predictive value of 75%, a negative predictive value of 90%, and an accuracy of 84% in the detection of vascular invasion on the basis of the findings of a stenosis.
Conclusion: On contrast-enhanced MR angiograms, findings of stenosis were sensitive and specific in the detection of arterial invasion. MR imaging evidence of partial or total encasement is highly specific in the detection of vascular invasion, while MR imaging evidence of a gap between the tumor and the vessels excludes an arterial invasion.
© RSNA, 2006
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Magnetic resonance (MR) imaging is now regarded as the optimal investigation in the evaluation of most primary musculoskeletal tumors (2). MR imaging is superior to other techniques in the staging of intraosseous tumor length and the assessment of the involvement of muscle compartments (3). The wide use of MR imaging and computed tomography (CT) has resulted in a decrease in the number of preoperative angiograms obtained. However, MR imaging does not enable delineation of the status of the neurovascular bundle substantially better than angiography, and, until now, angiography has remained the diagnostic method of choice in determining the precise relationship between the tumor and the neurovascular bundle (4).
The role of unenhanced MR angiography with two-dimensional time-of-flight and phase-contrast techniques in the preoperative staging of musculoskeletal neoplasms has been evaluated (5). In the study of Swan et al (5), MR angiography enabled adequate delineation of major vascular structures in the tumor bed. In recent years, unenhanced MR angiography has been supplanted by gadolinium-enhanced MR angiography because of the advantages of a shorter imaging time, reduced in-plane flow saturation effects, and decreased motion artifacts (6).
The purpose of our study was to prospectively evaluate the accuracy of contrast material–enhanced MR angiography in the evaluation of vascular invasion by bone and soft-tissue tumors, with surgery serving as the reference standard.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
A total of 22 tumors had developed in the bones: Of these tumors, eight were located in the distal femur; five, in the iliac bone; four, in the tibia; one, in the proximal femur; two, in the humerus; one, in the scapula; and one, in the fibula. Eight tumors developed in the soft tissue: Of these, three were located in the upper limb and the remaining five were located in the lower limb. In all patients, the diagnosis was confirmed by analyzing biopsy specimens before definitive surgery was performed.
Neoplasms included 12 osteosarcomas, four chondrosarcomas, three malignant fibrous histiocytomas, two synovial sarcomas, two benign cartilaginous tumors, one Ewing sarcoma, one rhabdomyosarcoma, one fibrosarcoma, one desmoid fibroma, one juvenile xanthogranuloma, one undifferentiated sarcoma, and one solitary metastasis. The metastatic tumor was solitary, and the patient was healthy. One patient had a cartilaginous tumor that was an osteochondroma of the proximal femur; this patient underwent surgery because the tumor was compressing the femoral vessels. The other benign cartilaginous tumor involved the acetabulum, and surgery was performed to avoid a fracture.
Six of the tumors were classified as local recurrences after previous surgical treatment. Twenty patients underwent neoadjuvant chemotherapy before surgery and after MR imaging. All patients underwent wide surgical resection of the tumor or amputation, with the exception of the four patients with benign tumors (cases 4, 14, 28, and 31). In these four patients, tumors were treated by means of marginal or contamined resection. In one patient with a large tumor in the pelvis (case 23), no curative resection was possible; in this patient, iliac venous ligation was performed at surgery. Our institutional review board approved this study, and informed consent was obtained from all patients.
MR Imaging and MR Angiography
All patients underwent MR imaging and MR angiography performed with a 1.5-T MR unit (Signa; GE Medical Systems, Milwaukee, Wis) shortly before surgical resection (mean interval between MR imaging and limb resection, 14 days; range, 1–60 days).
T1- and T2-weighted MR images were obtained in the transverse plane and at least one longitudinal plane (ie, coronal or sagittal plane) by using either a surface coil or a body coil. For T1-weighted imaging, a conventional spin-echo pulse sequence was used (repetition time msec/echo time msec, 400–600/8–16); for T2-weighted imaging, a fast spin-echo pulse sequence was used (3100–5300/80–100; echo train length, eight). All T2-weighted images were acquired with fat saturation and use of a fat frequency selective pulse or a fast short inversion time inversion-recovery sequence. Fat-saturated contrast-enhanced fast spin-echo T1-weighted images were obtained in all patients.
Sections were 5 mm thick, with a 1-mm intersection gap. The matrix size and field of view varied in accordance with the anatomic location and size of each lesion. For all MR imaging sequences, we used the smallest field of view that encompassed the region to be imaged, allowed a reasonable number of sections to be acquired, and provided an adequate signal-to-noise ratio. Superior and inferior presaturation pulses, no phase wrap, and no frequency wrap were selected to reduce artifacts.
MR angiography was performed immediately after MR imaging. MR angiography was performed with use of a three-dimensional fast spoiled gradient-echo sequence and intravenous injection of a gadolinium chelate at a rate of 1 mL/sec. MR angiograms were obtained by using fat saturation, subtraction, and maximum intensity projection in longitudinal planes. The field of view ranged from 18 to 44 cm, and it depended on the anatomic location. Section thickness was 2–3 mm.
The imaging parameters for typical MR angiography were as follows: repetition time msec/echo time msec/inversion time msec, 6.5/1.6/27.0. Typical imaging time was 50 seconds. MR angiography included arterial and venous phases. All conventional and MR angiographic examinations were performed within 1 week of each other (range, 1–6 days; mean, 3 days).
Contrast Material Administration
An injection of 0.2 mmol of gadopentetate dimeglumine (Dotarem; Guerbet, Villepinte, France) per kilogram of body weight was intravenously administered in all patients through an 18–22-gauge cannula in an antecubital vein. Contrast material was infused at a rate of 1 mL/sec by using an automatic injector. A delay of 15–20 seconds from the start of contrast material administration to the start of acquisition of the first contrast-enhanced images was used in all patients.
Angiography
In addition, 14 patients underwent digitalized angiography performed during the same week as MR angiography (range, 1–6 days after MR angiography; mean, 3 days after MR angiography). The other 16 patients did not undergo digitalized angiography because digitized angiography was performed, as per the requests of the orthopedists. Furthermore, this study was not designed as a comparative study of conventional angiography and MR angiography. All angiograms were obtained by the same radiologist (with 6 years of experience). In all patients, only arterial catheterization was performed. All arterial punctures were performed with the Seldinger technique. Digital subtraction was performed in one or more projections, with one run performed in the posteroanterior projection in all cases. A total of 140 mL of low-osmolar iodinated contrast material was used (Omnipaque 300; Nycomed, Oslo, Norway).
Interpretation of MR Images and MR Angiograms
All images were prospectively evaluated in consensus by two musculoskeletal radiologists (J.L.D. and A.F., with 14 and 5 years of experience, respectively, in MR imaging of musculoskeletal tumors). The readers evaluated MR images and MR angiograms obtained in the same patient at the same time; MR images were reviewed before MR angiograms. The readers were blinded to the results of conventional angiography and surgery, both of which were performed after MR imaging and MR angiography.
Anatomic MR images were evaluated for (a) the presence or absence of a fat plane between the vessels and tumor and (b) partial or total encasement of the vessels. The presence of a fat plane was defined as the visualization of a fat plane between the tumor and the main artery on all contiguous images. Partial vessel encasement was defined as a partial intratumoral course of the vessels (ie, arteries and veins), without complete encasement. Complete vessel encasement was defined as a complete intratumoral course of the vessels.
MR imaging scores were defined as follows: 0, fat; 1, no fat; 2, partial encasement; and 3, complete encasement. The presence of edema around vessels was also assessed on the fat-saturated fast spin-echo T2-weighted and fast short inversion time inversion-recovery transverse images.
MR angiographic images were evaluated for the presence of vascular displacement without stenosis, vascular displacement with stenosis, or occlusion. Maximum intensity projections were compared with the source images in all patients to exclude potential artifacts. MR angiographic findings were graded as follows: 0, free; 1, displacement; 2, stenosis of 0%–25%; 3, stenosis of 26%–50%; 4, stenosis of 51%–75%; 5, stenosis of 76%–99%; 6, occlusion. In cases in which stenosis was present, the length of the stenosis was measured in centimeters.
Interpretation of Conventional Angiograms
Angiograms were interpreted by an author (J.L.D., with 14 years of experience in the interpretation of musculoskeletal angiograms). The angiographer was blinded to the results of the MR examinations. Conventional angiographic findings were evaluated for displacement, stenosis, and occlusion. The angiographic findings were graded as follows: 0, free; 1, displacement; 2, stenosis of 0%–25%; 3, stenosis of 26%–50%; 4, stenosis of 51%–75%; 5, stenosis of 76%–99%; 6, occlusion. In the case of stenosis being present, the length of the stenosis was measured in centimeters.
Surgery
Surgery was performed by two surgeons (P.A. and B.T., with 11 and 32 years of experience, respectively, in musculoskeletal tumor surgery). For ethical reasons, surgeons knew the results of the MR imaging and MR angiographic examinations. The surgeons looked for the presence or absence of vessel encasement.
The features that were noted during the surgical procedure, especially the relationship between the vessels and the tumor, served as the reference standard. The detection of partial or complete encasement of a large artery was surgically confirmed by the presence of fixed adhesions (releasable adherence between the tumor and the vessel), which necessitated vascular resection with anastomosis. The term surgical positive means the artery could not be resected from the tumor, thereby necessitating further vascular surgery. The term surgical negative meant the artery could be freed from the tumor.
Statistical Analysis
Statistical analysis was performed by using surgical findings in surgically resected tumors as the reference standard. Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of MR imaging and MR angiographic findings were calculated separately for the presence of partial or complete encasement, stenosis, and occlusion. Statistical analyses were performed by using computer software (SPSS, version 10.0.7; SPSS, Chicago, Ill).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
MR Imaging Findings
Details of MR imaging findings are shown in Tables 1 and 2. At MR imaging, vessels appeared free, with a gap between the tumor and the vessel, in 12 cases. Contact between a vessel and a tumor without a gap was observed in 11 cases. In four cases, the vessels were included in the tumor, with partial encasement. Total vessel encasement was detected in four cases. Edema around vessels was observed in five cases. Only one of these cases (case 4) had vascular involvement with tumor encasement at surgery.
Angiographic Findings
In 14 cases, angiographic findings were available. Angiographic and MR angiographic results were in agreement in all but one case. In this case (case 19), MR angiography depicted a severe stenosis (>75%) and angiography showed a moderate stenosis (<25%); moderate stenosis was confirmed with surgical findings (releasable adhesion).
Accuracy of MR Imaging and MR Angiographic Findings
The accuracy of MR imaging and MR angiographic findings is shown in Tables 3 and 4. In one case (case 19), MR imaging showed that the vessels were partially encased; however, a releasable vessel adhesion was observed during surgery. In four cases (cases 6, 10, 21, and 27), MR imaging depicted contact (ie, no gap) between the vessels and the tumor; however, an unreleasable adhesion was observed during surgery. In these four cases, MR angiography depicted an arterial stenosis. In all but three patients with a gap between the tumor and the vessels at MR imaging, vessels were classified as free, without any adhesions at surgery. In the three cases with adhesions, the adhesions were surgically releasable.
|
|
Scores for MR imaging and MR angiography were tested for sensitivity, specificity, accuracy, and positive and negative predictive values; these findings are presented in Table 3. MR angiographic findings of a stenosis (ie, score of 2–6) had a sensitivity of 82%, specificity of 85%, positive predictive value of 75%, negative predictive value of 90%, and accuracy of 84% in the detection of vascular involvement.
MR imaging findings of partial or total encasement (score of 2 or 3) had a sensitivity of 64%, specificity of 95%, positive predictive value of 88%, negative predictive value of 83%, and accuracy of 84% in the detection of vascular involvement.
The combination of MR imaging findings of a partial or total encasement (ie, score of 2 or 3) and MR angiographic findings of a stenosis (ie, score of 2–6) had a sensitivity of 50%, specificity of 95%, positive predictive value of 86%, negative predictive value of 75%, and accuracy of 77% in the detection of vascular involvement.
The presence of MR imaging findings of partial or total encasement (ie, score of 2 or 3) or MR angiographic findings of stenosis (ie, score of 2–6) had a sensitivity of 79%, specificity of 100%, positive predictive value of 100%, negative predictive value of 85%, and accuracy of 90% in the detection of vascular involvement.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MR angiography allows noninvasive examination of the tumor bed vasculature, and it can be performed at the end of the MR imaging staging examination in most patients with musculoskeletal tumors. In two previous studies (5,7), two-dimensional time-of-flight unenhanced MR angiography was used.
Lang et al (7) examined 13 patients with primary musculoskeletal neoplasms involving a limb. Two-dimensional time-of-flight MR angiography was performed, and two-dimensional maximum intensity projections and three-dimensional displays were analyzed. This study did not include conventional angiography. MR angiography depicted vascular displacement when the tumor encroached on the major arteries or veins and showed an abrupt area of vascular stenosis or vascular tapering when the tumor encased the vessels (7).
Swan et al (5) used two-dimensional time-of-flight MR angiography during staging of musculoskeletal tumors in 23 patients. In this series, MR angiography adequately depicted major vascular structures in the tumor bed. Abnormal vasculature findings—such as stenoses, occlusions, and neovascularity—were uncommon. According to Swan et al (5), encasement and stenosis may be relatively uncommon in tumor beds, with the tumor "pushing" the vasculature away instead. False-positive stenoses were observed, and may have been related to an overestimation of the stenosis with the time-of-flight technique. Small feeder vessels were not reliably identified because MR angiograms did not show the neovascular blush. This problem is not crucial, however, as the local control of flow in the main vessels is of greatest concern during surgery performed in limb tumors.
Advances in gadolinium-enhanced three-dimensional MR angiography have facilitated rapid imaging of the arterial tree during the arterial phase with use of a gadolinium chelate bolus. These techniques have proved to be robust, accurate, and reproducible in clinical practice and have essentially replaced time-of-flight MR angiography in many areas (6). We used a gadolinium-enhanced three-dimensional MR angiographic technique, with subtraction of unenhanced and gadolinium-enhanced images to minimize background signal.
Because intravascular signal intensity depends entirely on gadolinium-induced T1 shortening, there is no susceptibility to in-plane saturation. This enables exploitation of the three-dimensional acquisition and coronal orientation of the imaging volume for maximum anatomic coverage with any combination of field of view and section thickness (10).
All MR angiographic examinations were successful and allowed us to evaluate the relationship between the tumor and the major vessels. We did not observe differences in the technical quality of the MR angiographic examinations between the various tumor locations.
Our patients were all referred and examined before surgery. Surgical findings showed that three groups of similar size were actually studied: Free vessels were studied in 10 cases; releasable adhesions, in 10 cases; and nonreleasable adhesions, in 11 cases. The absence of arterial occlusion in our patients was not surprising, as bone and soft-tissue tumors usually displace and encase the vessels (1).
Predictive Value of MR Angiographic Findings in Our Study
Eight cases were classified as being free of tumor at MR angiography. This finding had an excellent negative predictive value (100%) for arterial invasion, with six of eight cases being graded as free of tumor and releasable adhesions being diagnosed in the remaining two cases at surgery.
When the vessels were displaced without stenosis at MR angiography, the risk of encroachment and encasement of the vessels was low, but we observed two false-negative results (ie, fixed adhesions in the first case and encasement in the second case). In both cases, however, MR imaging depicted a partial encasement, which was confirmed during the surgical procedure. The association between MR angiography and MR imaging may enable physicians to reliably diagnose an encroachment or an encasement of the bundle vessel close to the tumor.
In our study, we observed two false-positive results of MR angiography. In these two patients, MR angiography depicted a stenosis that was not observed during the surgical procedure. MR imaging also depicted an encroachment in one of these two patients. In this patient, angiographic findings were not in agreement with MR imaging findings, and MR angiography showed only a moderate stenosis. Overestimation of the stenosis with MR angiography could be related to flow turbulences.
Predictive Value of MR Imaging Findings in Our Study
All but three cases with a gap between the tumor and the vessels at MR imaging were classified as free, without any adhesions at surgery. The three vessels with adhesions could be released. Hence, our results show that observation of a completely preserved soft-tissue interface (fat plane) on T1-weighted images, even in the presence of edema around the vessels, has a negative predictive value of 100% for arterial invasion.
It was determined that among the four patients with partial encasement at MR imaging, three had arterial invasion at surgery. All four patients with MR imaging findings of vessel encasement also had encasement at surgery. The overall sensitivity of MR imaging for arterial invasion criteria (ie, partial or complete encasement) was moderate (0.64).
The presence of edema around vessels was not a reliable predictive sign of arterial invasion in our series. Peritumoral edema or reactive change may indicate a vascular or inflammatory response with or without tumor cell infiltration. The reactive zone around a sarcoma commonly is removed en bloc with the tumor during limb salvage surgery.
Limitations
Because of the relatively few cases of arterial invasion, caution must be exercised in interpreting the absolute statistical values obtained for each MR imaging and MR angiographic sign. In addition, selection bias may have been introduced into this study by our inclusion of only those subjects in whom a tumor was likely to be resected at surgery.
There are potential pitfalls in the interpretation of MR imaging signs of arterial invasion (ie, the presence of a fat plane between the arterial wall and the tumor). Oblique vessel course and partial volume averaging may cause the arterial border to have an ill-defined appearance on transverse MR images, which leads to a false-positive finding of possible arterial invasion. In case 14, MR imaging showed no fat plane between the iliac artery and the tumor, whereas MR angiography showed an arterial displacement without stenosis. The vessels were free without any adhesion at surgery.
Response to chemotherapy was not evaluated in our study because we obtained MR angiograms in only three patients before neoadjuvant chemotherapy was started. In the remaining patients, MR angiography was performed before surgery and after chemotherapy, or the patients did not receive neoadjuvant chemotherapy. Fujii et al (11) and Kunisada et al (12) demonstrated the utility of angiography in the assessment of preoperative chemotherapy in musculoskeletal sarcomas (11). If MR imaging is not a good technique in the assessment of the efficiency of preoperative chemotherapy, MR angiography may yield better results. Lang et al (7) demonstrated the utility of MR angiography in assessing the response to chemotherapy in patients with osteogenic sarcoma.
The combination of MR imaging and contrast-enhanced MR angiography is highly sensitive and specific in the detection of arterial invasion in patients with musculoskeletal limb tumors. On contrast-enhanced MR angiograms, findings of stenosis were sensitive and specific in the detection of arterial invasion.
By the current standards, most patients with musculoskeletal neoplasms will be examined with MR imaging because it is superior to other imaging techniques in enabling radiologists to define and stage the extent of the lesion. MR angiography can be appended to this routine examination, with only a small time increase (ie, 2 minutes) by using the described imaging sequence.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, A.F., P.A., J.L.D.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, A.F., P.A.; clinical studies, P.A., B.T., A.C., J.L.D.; statistical analysis, A.F., P.A., J.L.D.; and manuscript editing, A.F.
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Related Article
This article has been cited by other articles:
|
|
 |
|
  M. V. Marx, A. Feydy, and J.-L. Drape Arterial Anatomy in Feydy et al Radiology, March 1, 2007; 242(3): 949 - 950. [Full Text] [PDF]
|
|
|
|
|
|
M. Karcaaltincaba and A. Feydy Multidetector CT Angiography for Vascular Invasion of Musculoskeletal Tumors Radiology, February 1, 2007; 242(2): 634 - 635. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
