Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

DOI: 10.1148/radiol.2232010406
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
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 Bodner, G.
Right arrow Articles by Jaschke, W. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bodner, G.
Right arrow Articles by Jaschke, W. R.
(Radiology 2002;223:410-416.)
© RSNA, 2002

Musculoskeletal Imaging

Differentiation of Malignant and Benign Musculoskeletal Tumors: Combined Color and Power Doppler US and Spectral Wave Analysis1

Gerd Bodner, MD, Michael F. H. Schocke, MD, Franz Rachbauer, MD, Klaus Seppi, MD, Siegfried Peer, MD, Anke Fierlinger, MD, Tarek Sununu, MD and Werner R. Jaschke, MD

1 From the Departments of Radiology (G.B., M.F.H.S., S.P., W.R.J.), Orthopedics (F.R., A.F., T.S.), and Neurology (K.S.), University Hospital of Innsbruck, Anichstrasse 35, 6020 Innsbruck, Austria. Received February 5, 2001; revision requested March 26; revision received August 6; accepted September 28. Address correspondence to G.B. (e-mail: gerd.bodner@uibk.ac.at).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
PURPOSE: To assess the use of combined color Doppler ultrasonography (US), power Doppler US, and spectral wave analysis (SWA) in differentiating malignant and benign musculoskeletal tumors.

MATERIALS AND METHODS: Seventy-nine musculoskeletal tumors (34 malignant, 45 benign) were examined with color and power Doppler US and SWA. Two radiologists independently assessed US images and SWA findings. Echotexture and vessel characteristics such as stenoses, occlusions, loops, shunts, trifurcations, vascular pattern, and resistive index were evaluated. All tumors were subject to US-guided or open biopsy for histologic correlation.

RESULTS: Combined color and power Doppler US and SWA revealed four major (stenosis, occlusion, trifurcation, vascular pattern) and three minor (shunt, self loop, resistive index) vessel characteristics, which proved helpful in differentiating benign from malignant lesions. Echotexture showed moderate sensitivity (82% [28 of 34 tumors]) and low specificity (38% [17 of 45 tumors]). When comparing several combinations of vessel characteristics, a combination of any two major characteristics demonstrated the best results (sensitivity, 94% [33 of 39 tumors]; specificity, 93% [three of 45 tumors]). Combining more than two characteristics resulted in lower sensitivity.

CONCLUSION: Combined color and power Doppler US and SWA may enable assessment of vascular architecture and altered flow of musculoskeletal tumors. Vascular architecture analysis enables differentiation of benign and malignant lesions and evaluation of musculoskeletal tumors.

© RSNA, 2002

Index terms: Muscles, neoplasms • Ultrasound (US), Doppler studies, 40.12984, 40.12985 • Ultrasound (US), power Doppler studies, 40.12984, 40.12985 • Ultrasound (US), tissue characterization, 40.12984, 40.12985


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Various imaging techniques have been used to assess musculoskeletal tumors. Evaluation with gray-scale ultrasonography (US) provides rapid, relatively inexpensive, and noninvasive assessment of lesion morphology. This technique is a helpful tool for lesion localization and guided needle biopsy but is not reliable enough to characterize the nature of a lesion (1–3). Magnetic resonance (MR) imaging has been established for detection of tumors; it provides information about morphologic characteristics, signal intensity characteristics, and rate and pattern of vascular enhancement (4) but cannot reliably enable distinction of benign from malignant lesions (5). Although angiography can depict vascular abnormalities such as neovasculature, regional blood flow alteration, and obstruction of draining veins, it does not permit accurate differentiation of benign lesions from malignant ones with high accuracy (6).

Color Doppler US permits high-spatial-resolution imaging of small vessels (7) and has demonstrated a high capability in depicting and characterizing vascular soft-tissue anomalies (8,9), and power Doppler US accurately depicts tumor vascularity (10). Nevertheless, whether color or power Doppler US or spectral wave analysis (SWA) can enable differentiation between benign and malignant musculoskeletal tumors is not clear (10–19).

The purpose of our study was to assess the use of combined color and power Doppler US and SWA in differentiating between malignant and benign musculoskeletal tumors.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Between April 1997 and March 2000, 145 patients with a clinically palpable mass were referred to our department for suspected musculoskeletal tumor and were included in a prospective study using combined gray-scale and color and power Doppler US and SWA for evaluation. The main inclusion criterion was having at least five tumor vessels, regardless of tumor size. Seventy-nine musculoskeletal tumors that showed at least five tumor vessels in 79 patients (mean age, 45 years; range, 12–70 years; 38 female and 41 male patients) were available for evaluation. Sixty-six patients were excluded because they had no tumors or low-vascularized tumors such as simple cysts (n = 30), hematomas (n = 14), or small lipomas (n = 22). While simple cysts and hematomas were mainly deep seated, a majority of small lipomas were subcutaneous. In these lesions, no vessels could be visualized with color Doppler US. These tumors were either surgically removed, without evidence of malignancy, or followed up with US and MR imaging.

Patients gave written consent for US examination and biopsy. Investigative and interventional procedures were performed on the basis of the guidelines of the Helsinki Declaration (20).

Gray-scale US (HDI 3000 or 5000; Advanced Technology Laboratories, Bothell, Wash) in combination with color and power Doppler US and SWA were prospectively performed by one of two investigators (G.B., M.F.H.S.). The first investigator (G.B.) had more than 6 and the second (M.F.H.S.) more than 2 years of experience in musculoskeletal US. Images from the whole US examination were stored on foil prints and electronically. To determine interobserver agreement, images from gray-scale and color and power Doppler US and SWA in at least five vessels were reviewed by the other investigator, who was blinded to patient data and final histologic diagnosis.

Diagnostic gray-scale US was performed to assess size, location, and internal echotexture of the tumor. At gray-scale US, tumor echotexture was classified as hyperechoic or hypoechoic with reference to surrounding muscle.

The tumor was classified as inhomogeneous or homogeneous. The vascular pattern of each tumor was assessed by using color and power Doppler US. SWA was performed in at least five arterial tumor vessels.

The following color and power Doppler US and SWA criteria were applied:

1. Self or true loop was defined when a self-connective tumor vessel with side branches was visible during color and power Doppler US (21). The loop was scanned in a second plane to confirm the self-vascular connection.

2. Trifurcation was defined when the original vessel was divided into three vessels, with branches originating from the same point (21).

3. Stenosis was diagnosed primarily with color Doppler US while visually searching for aliasing phenomena (22). When a tumor vessel segment demonstrated caliber reduction, as indicated by aliasing phenomena, SWA was performed proximal to, at, and distal to the caliber change. Increase in maximum peak flow velocity in the caliber reduction by more than 100% was considered stenosis (23).

4. An occluded vessel was identified with monophasic waveform and absence of diastolic flow, indicating occlusion in a distal part of the same vessel (23,24).

5. Arteriovenous shunt was diagnosed when a low resistive index (RI) of less than 0.5 was found (9).

6. RI was calculated as maximum systolic velocity - minimum diastolic velocity/maximum systolic velocity. RI was measured in at least five vessels. We determined the ratio of lowest to highest value (RImin/max), where min = minimum and max = maximum.

7. Two vascular patterns were distinguished by using color and power Doppler US: (a) a hierarchic vascular tree, with continuously diminishing vessel size toward the lesion periphery, representing normal vessel architecture (Fig 1); and (b) an anarchic vascular pattern, as described by Less et al (21), consisting of caliber changes, loops, and trifurcations (Fig 2).



View larger version (202K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Neurofibrosarcoma (4-6 cm) in right gluteal area in a 55-year-old patient. Three-dimensional power Doppler US image shows anarchic vessel architecture with multiple loops (solid arrows) and caliber changes (open arrows).

 


View larger version (203K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. Myositis proliferans in right thigh of a 16-year-old patient. Three-dimensional power Doppler US image demonstrates prominent tumor vessels (arrows) diminishing in size to smaller vessels (arrowheads) at the tumor periphery.

 
Gray-scale and color and power Doppler US examinations were performed (HDI 3000 or 5000; Advanced Technology Laboratories, Bothell, Wash) by using linear arrays at 4–7 and 5–12 MHz. The transducer was chosen according to the size and location of the tumor (lesions more than 3 cm below skin level were examined with the 4–7-MHz array, and lesions within 3 cm below skin level were investigated with the 5–12-MHz array). Color Doppler US settings were adjusted for investigation of low- or high-flow tumor vessels. Pulse repetition frequency was set at 800 Hz to detect low-flow vessels (ie, occlusion, loop) and at 2,500 Hz to detect high-flow vessels (ie, stenosis, arteriovenous shunt). Waves in a wave train were averaged for measurements. We used power Doppler US to assess the vascular architecture of the tumor, with the pulse-repetition frequency set at 1,000 Hz. The Doppler angle was set below 60°. Three-dimensional power Doppler US was performed to better demonstrate the vascular pattern of the tumor.

All tumors (n = 79) were subjected to US-guided core-needle biopsy, without any adverse effects. Because of inconclusive biopsy findings, only two patients underwent surgical biopsy. A total of 42 tumors were resected, and further histopathologic evaluation confirmed biopsy findings in all cases; 37 patients did not undergo surgery because of the benign nature of the tumors or the advanced stage of disease.

Histologic examination revealed 34 malignant tumors, including fibrous (n = 3), osseous (n = 3), metastatic (n = 11), vascular (n = 2), lipomatous (n = 5), neural (n = 2), smooth-muscle (n = 5), synovial (n = 2), and unclassified (n = 1) tumors. Forty-five benign tumors included fibrous (n = 13), osseous (n = 4), lipomatous (n = 8), neural (n = 5), vascular (n = 10), and synovial (n = 1) tumors, as well as infectious lesions (n = 4) (Tables 1 and 2). Two intraarticular synovial processes were included in the study, since they arose into the suprapatellar pouch, with extensions in the quadriceps femoris muscle, which made them visible and palpable as tumors of the distal femur. Patients with bone tumors were included in the study because they had a large, dominant, soft-tissue component at presentation. All soft-tissue metastases were true soft-tissue metastases. The patient with tuberculosis and the two patients with abscesses were included, since there were no clinical signs of infection (cold abscesses).


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

  		TABLE 1. Final Diagnosis, Number, and Diameter of Malignant Tumors

 

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

  		TABLE 2. Final Diagnosis, Number, and Diameter of Benign Tumors

 
Computer software (SPSS 10.0 for Windows; SPSS, Chicago, Ill) was used for statistical analysis. Interobserver reliability was measured by using a generalized {kappa} statistic. Similar to a correlation coefficient, {kappa} varies from -1.0 (complete disagreement) to 0 (chance agreement) to +1.0 (perfect agreement). Strength of agreement was labeled as poor ({kappa} < 0), slight ({kappa} = 0–0.20), fair ({kappa} = 0.21–0.40), moderate ({kappa} = 0.41–0.60), substantial ({kappa} = 0.61–0.80), or excellent ({kappa} = 0.81–1.00), as suggested in the literature (25–28). Whenever ratings differed between observers, the incorrect diagnosis was used for further statistical analysis.

The unpaired t test was performed to compare RIs between benign and malignant musculoskeletal tumors. Distribution of vessel characteristics in benign and malignant musculoskeletal tumors was analyzed with the {chi}2 test for categoric variables. Sensitivity, positive predictive value (PPV), specificity, and negative predictive value (NPV) were calculated for each putative vessel characteristic (absent vs present) with respect to histopathologic diagnosis (29). Sensitivity refers to the probability that color and power Doppler US and SWA reveal malignant tumor when histopathologic findings indicate malignant tumor. PPV refers to the probability of a tumor being malignant when US depicts a malignant tumor. Specificity refers to the probability that US depicts a benign lesion when histopathologic examination shows a benign tumor. NPV refers to the probability of having a benign tumor when US demonstrates a benign lesion.

Vessel characteristics were then evaluated with multivariate analysis by using stepwise logistic regression to identify the set of criteria that would best predict the diagnosis of malignancy. Vessel characteristics representing the best model to indicate malignancy in the multivariate logistic regression analysis were defined as major variables; those not included in the model were defined as minor variables (4,17,18,30).

Sensitivity and specificity of RImin/max were assessed with a receiver operating characteristic curve, which plots sensitivity versus specificity for every possible cutoff point (31). The optimal cutoff ratio, approximately 0.67, was determined visually on the receiver operating characteristic curve. This cutoff value represents the best ratio of minimal and maximal RI for discrimination of benign versus malignant tumors. According to the calculated cutoff value, we divided all tumors into two groups with an RImin/max higher or lower than 0.67 for further categoric testing.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Interobserver reliability analysis showed moderate agreement for imaging findings such as echotexture ({kappa} = 0.58) and homogeneity ({kappa} = 0.42), good agreement for loops ({kappa} = 0.66), shunts ({kappa} = 0.78), occlusions ({kappa} = 0.78), vascular pattern ({kappa} = 0.80), and trifurcations ({kappa} = 0.80), and excellent agreement for stenoses ({kappa} = 0.85).

Evaluation of echotexture provided a sensitivity of 82% (28 of 34 tumors), a PPV of 50% (28 of 56 tumors), a specificity of 38% (29 of 56 tumors), and an NPV of 74% (17 of 23 tumors) for detection of malignancy in musculoskeletal tumors. In contrast, homogeneity had a sensitivity of 85% (29 of 34 tumors), a PPV of 52% (29 of 56 tumors), a specificity of 40% (18 of 45 tumors), and an NPV of 78% (18 of 23 tumors).

The distribution of RI within the tumor tissue, expressed by the ratio of RImin to RImax (RImin/max) for all vessels measured, differed significantly between malignant (mean RImin/max, 0.50 ± 0.19 [SD]) and benign tumors (mean RImin/max, 0.79 ± 0.12; P < .001). Sensitivity, PPV, specificity, and NPV for each vessel characteristic are listed in Table 3. Since all vessel characteristics had a P value less than .05 with the {chi}2 test (Table 3), all were included in subsequent multivariate analysis. Stepwise logistic regression analysis revealed a combination of various vessel characteristics, which included stenoses, occlusions, trifurcations, and anarchic vascular pattern as the best model (P < .001), defining them as major variables, in contrast with shunts, self loops, and RImin/max, which were minor variables. The diagnostic value of different combinations of vessel characteristics such as sensitivity, PPV, specificity, and NPV for any one, two, three, and four of all vessel characteristics can be seen in Table 3. Combinations of major and minor variables were separately tested for sensitivity, PPV, specificity, and NPV, as presented in Table 3. The combinations of any two major characteristics appeared to be the best compromise between sensitivity and specificity. In addition, PPVs and NPVs for these combinations were excellent. In comparison, the combinations of any two minor characteristics or any two random characteristics showed poorer predictive values.


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

  		TABLE 3. Diagnostic Accuracy of Combined Color and Power Doppler US and SWA Findings in Malignant and Benign Musculoskeletal Tumors

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Despite the fact that soft tissue and bone compose almost two-thirds the mass of the human body, sarcomas are uncommon tumors. Benign neoplasms of the soft tissues, in contrast, are commonplace and rarely of consequence. Because of the relative rarity of sarcomas, as compared with benign soft-tissue tumors, patients and clinicians frequently fail to appreciate the clinical importance of an enlarging soft-tissue mass, and diagnosis is commonly obtained only after major delay (32). Benign and malignant soft-tissue tumors commonly present as a painless mass. To our knowledge, there are no reliable findings at physical examination to determine whether a musculoskeletal mass is benign or malignant. Benign soft-tissue tumors far outnumber their malignant counterparts. Malignant musculoskeletal tumors account for fewer than 1% of all neoplasms, whereas benign musculoskeletal masses such as cysts and lipomas occur frequently (32). Nevertheless, virtually all soft-tissue masses that are painful, appear suddenly, or are over 5 cm in diameter, as well as any new, enlarging, or symptomatic lesions, need to be investigated with scrutiny. To our knowledge, none of the existing imaging methods, such as MR imaging, computed tomography (CT), and US, reliably enable distinction between malignant and benign entities (5,33). Therefore, final diagnosis must be confined to histologic results of core-needle or open biopsy (1,2). Since clinical and imaging methods are so nonspecific, the routine diagnostic procedures for evaluating musculoskeletal tumors until final diagnosis take time and require intensive workup. Consequently, if imaging techniques were to permit accurate evaluation of tumor entity, the diagnostic procedure would be shortened.

Our study is based on US evaluation of musculoskeletal tumors, since US has excellent spatial resolution, is widely available, and is low in cost. In previous studies (1,3,8–15), as described earlier, US examination has been limited by the lack of uniform criteria for malignancy. We therefore propose US criteria that may help differentiate between benign and malignant tumors.

Gray-scale US has a high sensitivity for detecting tumors but has not proven useful in differentiating between benign and malignant conditions (1–3). The contribution of gray-scale US to the final diagnosis in our study was minor. Echotexture and homogeneity can be used to differentiate between a benign or malignant tumor with specificities of 75.6% or 40.0%, respectively.

Unfortunately, when using echotexture and homogeneity to distinguish between benign and malignant tumors, 84.6% and 49.5% of the patients with a diagnosis of malignant tumor had a benign one.

Histologic studies show that the vascular heterogeneity of tumors depends on their degree of malignancy. Progressive growth of malignant tumors requires a newly formed blood supply induced by neoangiogenesis, which primarily starts in venules and subsequently involves the arteries (21,34–36). Malignant tumor vessels are histologically characterized by lack of the muscular layer and by irregular contours. They commonly form a heterogeneous reticular network induced by neovascularization. In these chaotic vascular architectures of malignant tumors, occlusions, stenoses, arteriovenous shunts, loops, and trifurcations occur (21). The vascular abnormalities of vessels in malignant tumors result in heterogeneous regional flow patterns (35).

Color and power Doppler US permit detection of vessels as small as 0.2 mm in diameter (7), which is why they are good tools for investigating the vascular architecture of tumors. Several authors (3,8,10,12) have demonstrated the usefulness of color Doppler US for imaging vascular branches in tumors. Accurate visualization of tumor vessels, as permitted by the technical development of color and power Doppler US, has resulted in several attempts to define US criteria for malignancy. Schroeder et al (10) point out that sonomorphologic analysis of tumor vessels appears to be much more reliable than quantitative parameters such as flow velocities or calculated indices. The mean systolic peak flow velocity is said to be higher in malignant tumors, as compared with benign tumors (13,37,38). Paltiel et al (9) and Dubois et al (39) have concluded that high vessel density and high peak arterial Doppler shift can distinguish hemangiomas from other soft-tissue masses. High-frequency Doppler shift in malignant entities has been detected in two studies (11,15). These findings show that color and power Doppler US and SWA can depict vascular irregularities in malignant tumors, whereas the criteria for malignancy, as reported in the literature, vary widely. Other authors (3,14,18,19) controversially discuss or even deny the ability of color and power Doppler US and SWA to enable distinction of benign from malignant lesions.

Our findings show that combined color and power Doppler US and SWA can differentiate between benign and malignant tumors and demonstrate occlusions, stenoses, shunts, trifurcations, and loops. Self loops and trifurcations are detected mainly with power Doppler US, which provides accurate depiction of reticular vessel architecture (Figs 1, 3). In addition, SWA reveals a typical pattern of self loops, characterized by a truncated systolic peak associated with a relatively high diastolic phase (Fig 3b). Corresponding with the visualization of trifurcations and self loops, occlusions, stenoses, and shunts can be primarily demarcated with the help of color and power Doppler US followed with SWA (Fig 4). Consequently, vessel characteristics are verified with color and power Doppler US and SWA so that the origin of signals is reliably identified.



View larger version (164K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3a. Liposarcoma in right inguinal region of a 61-year-old patient. (a) Longitudinal color Doppler US image shows tumor vessel at tumor margin, with a connection to itself corresponding to a self loop (arrows). Images obtained in a second plane (not shown) confirmed self connection. (b) SWA of self loop shows peak systolic flow of 40.4 cm/sec (curved arrow) and high diastolic flow of 20.2 cm/sec (straight arrow), with a resulting RI of 0.5.

 


View larger version (123K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3b. Liposarcoma in right inguinal region of a 61-year-old patient. (a) Longitudinal color Doppler US image shows tumor vessel at tumor margin, with a connection to itself corresponding to a self loop (arrows). Images obtained in a second plane (not shown) confirmed self connection. (b) SWA of self loop shows peak systolic flow of 40.4 cm/sec (curved arrow) and high diastolic flow of 20.2 cm/sec (straight arrow), with a resulting RI of 0.5.

 


View larger version (148K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4a. Color Doppler US images obtained in a 71-year-old patient. (a) Longitudinal scan shows chondrosarcoma in left femur, with SWA in tumor vessel (curved arrow), with maximal systolic flow velocity increase of 110 cm/sec (straight arrow), representing a stenosed tumor vessel. Flow velocity prior to stenosed vessel segment was 45 cm/sec (image not shown). (b) Transverse scan shows chondrosarcoma. SWA of occluded tumor vessel shows reduced systolic flow of 9 cm/sec (solid arrow) and no diastolic flow (open arrow), suggesting that this vessel is distally occluded. (c) Longitudinal scan shows chondrosarcoma in left femur, with SWA of arteriovenous shunt showing a delayed systolic peak (left arrow) and high diastolic flow (right arrow), with resulting RI of 0.35, suggesting direct connection from arterial to venous tumor vessel.

 


View larger version (170K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4b. Color Doppler US images obtained in a 71-year-old patient. (a) Longitudinal scan shows chondrosarcoma in left femur, with SWA in tumor vessel (curved arrow), with maximal systolic flow velocity increase of 110 cm/sec (straight arrow), representing a stenosed tumor vessel. Flow velocity prior to stenosed vessel segment was 45 cm/sec (image not shown). (b) Transverse scan shows chondrosarcoma. SWA of occluded tumor vessel shows reduced systolic flow of 9 cm/sec (solid arrow) and no diastolic flow (open arrow), suggesting that this vessel is distally occluded. (c) Longitudinal scan shows chondrosarcoma in left femur, with SWA of arteriovenous shunt showing a delayed systolic peak (left arrow) and high diastolic flow (right arrow), with resulting RI of 0.35, suggesting direct connection from arterial to venous tumor vessel.

 


View larger version (52K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4c. Color Doppler US images obtained in a 71-year-old patient. (a) Longitudinal scan shows chondrosarcoma in left femur, with SWA in tumor vessel (curved arrow), with maximal systolic flow velocity increase of 110 cm/sec (straight arrow), representing a stenosed tumor vessel. Flow velocity prior to stenosed vessel segment was 45 cm/sec (image not shown). (b) Transverse scan shows chondrosarcoma. SWA of occluded tumor vessel shows reduced systolic flow of 9 cm/sec (solid arrow) and no diastolic flow (open arrow), suggesting that this vessel is distally occluded. (c) Longitudinal scan shows chondrosarcoma in left femur, with SWA of arteriovenous shunt showing a delayed systolic peak (left arrow) and high diastolic flow (right arrow), with resulting RI of 0.35, suggesting direct connection from arterial to venous tumor vessel.

 
According to our results, malignancy is best represented by the combination of occlusions, stenoses, trifurcations, and vascular pattern. Logistic regression analysis makes these major characteristics for malignancy, whereas RI, loops, and shunts constitute minor characteristics. The different effect of minor vascular characteristics in predicting malignancy is reflected by the vascular pattern of benign vascular tumors, which also contain reticular vessels such as self loops and shunts, as seen in arteriovenous malformations contributing to abnormal RI values. As a consequence, these three characteristics are not typical for malignant tumors. The major characteristics, such as trifurcations and vascular pattern, have both a high sensitivity and a high specificity in comparison with occlusions and stenoses, which are hampered by lower sensitivity. Combinations of several characteristics increase sensitivity but decrease specificity. Combinations of more than two vessel characteristics provide increasing specificity and PPV at the price of lower sensitivity and NPV.

Combined color Doppler US and SWA permit noninvasive visualization of abnormal vessel architecture, which is commonly found in malignant tumors, as reported in several histologic studies (21,32,40). In agreement with histopathologic findings of malignant and benign tumors, we introduce the concept of major malignancy-related vessel characteristics.

Although this study’s findings show the usefulness of color and power Doppler US and SWA in differentiating benign from malignant tumors, some limitations need to be addressed. Malignant tumors less than 1.5 cm in diameter do not induce US-detectable malignant neovascularity (17), which is why these tumors appear nonspecific when using color Doppler US and SWA. Furthermore, completely necrotic tumors also show a lack of vascularization and thus are not accessible for color and power Doppler US.

Biopsy was performed on all musculoskeletal tumors in our series with the help of gray-scale US. Only in two patients were biopsies inconclusive, so that open biopsy was necessary. With reference to the literature (1,2), our score in US-guided core-needle biopsy was excellent.

In conclusion, our study shows that combined color and power Doppler US and SWA permit reproducible visualization of abnormal vascular architecture and enable differentiation between benign and malignant tumors. It enables identification of so-called major and minor vessel characteristics, which are essential for further assessment of tumor vascularization. The best compromise with respect to sensitivity, specificity, PPV, and NPV in differentiating the dignity of musculoskeletal tumors is achieved when combining any two major vessel characteristics. Consequently, combined color and power Doppler US and SWA are most helpful for primary imaging of musculoskeletal tumors.


   FOOTNOTES
 
Abbreviations: NPV = negative predictive value, PPV = positive predictive value, RI = resistive index, SWA = spectral wave analysis

Author contributions: Guarantors of integrity of entire study, G.B., M.F.H.S., K.S.; study concepts, G.B., M.F.H.S., S.P.; study design, G.B., M.F.H.S., K.S., W.R.J.; literature research, G.B., M.F.H.S.; clinical studies, G.B., M.F.H.S., S.P.; data acquisition, G.B., M.F.H.S.; data analysis/interpretation, all authors; statistical analysis, G.B., M.F.H.S., K.S.; manuscript preparation, G.B., M.F.H.S., S.P., K.S., W.R.J.; manuscript definition of intellectual content, G.B., M.F.H.S., A.F., W.R.J.; manuscript editing and revision/review, G.B., M.F.H.S., S.P., W.R.J.; manuscript final version approval, G.B., M.F.H.S., W.R.J.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Alexander AA, Nazarian LN, Feld RI. Superficial soft-tissue masses suggestive of recurrent malignancy: sonographic localization and biopsy. AJR Am J Roentgenol 1997; 169:1449-1451.[Abstract/Free Full Text]
  2. Yao L, Nelson SD, Seeger LL, Eckardt JJ, Eilber FR. Primary musculoskeletal neoplasms: effectiveness of core-needle biopsy. Radiology 1999; 212:682-686.[Abstract/Free Full Text]
  3. Lagalla R, Iovane A, Caruso G, Lo Bello M, Derchi LE. Color Doppler ultrasonography of soft-tissue masses. Acta Radiol 1998; 39:421-426.[Medline]
  4. Ma LD, Frassica FJ, McCarthy EF, Bluemke DA, Zerhouni EA. Benign and malignant musculoskeletal masses: MR imaging differentiation with rim-to-center differential enhancement ratios. Radiology 1997; 202:739-744.[Abstract/Free Full Text]
  5. Kransdorf MJ, Murphey MD. Radiologic evaluation of soft-tissue masses: a current perspective. AJR Am J Roentgenol 2000; 175:575-587.[Free Full Text]
  6. Viamonte M, Jr, Roen S, LePage J. Nonspecificity of abnormal vascularity in the angiographic diagnosis of malignant neoplasms. Radiology 1973; 106:59-63.[Medline]
  7. Voelckel S, Bodner G, Voelckel W, Stadlwieser C, De Koekkoek P, Springer P. Doppler ultrasound determination of vascular resistance in arteriovenous shunts of the finger tip. Ultraschall Med 1998; 19:181-186.[Medline]
  8. Trop I, Dubois J, Guibaud L, et al. Soft–tissue venous malformations in pediatric and young adult patients: diagnosis with Doppler ultrasonography. Radiology 1999; 212:841-845.[Abstract/Free Full Text]
  9. Paltiel HJ, Burrows PE, Kozakewich HP, Zurakowski D, Mulliken JB. Soft-tissue vascular anomalies: utility of ultrasonography for diagnosis. Radiology 2000; 214:747-754.[Abstract/Free Full Text]
  10. Schroeder RJ, Maeurer J, Vogl TJ, et al. D-galactose-based signal-enhanced color Doppler sonography of breast tumors and tumorlike lesions. Invest Radiol 1999; 34:109-115.[CrossRef][Medline]
  11. Taylor KJ, Ramos I, Carter D, Morse SS, Snower D, Fortune K. Correlation of Doppler ultrasonography tumor signals with neovascular morphologic features. Radiology 1988; 166:57-62.[Abstract/Free Full Text]
  12. Taylor GA, Perlman EJ, Scherer LR, Gearhart JP, Leventhal BG, Wiley J. Vascularity of tumors in children: evaluation with color Doppler imaging. AJR Am J Roentgenol 1991; 157:1267-1271.[Abstract/Free Full Text]
  13. Dock W, Grabenwoger F, Metz V, Eibenberger K, Farres MT. Tumor vascularization: assessment with duplex sonography. Radiology 1991; 181:241-244.[Abstract/Free Full Text]
  14. Latifi HR, Siegel MJ. Color Doppler flow imaging of pediatric soft tissue masses. J Ultrasound Med 1994; 13:165-169.[Abstract]
  15. Van Campenhout I, Patriquin H. Malignant microvasculature in abdominal tumors in children: detection with Doppler ultrasonography. Radiology 1992; 183:445-448.[Abstract/Free Full Text]
  16. Meyerowitz CB, Fleischer AC, Pickens DR, et al. Quantification of tumor vascularity and flow with amplitude color Doppler sonography in an experimental model: preliminary results. J Ultrasound Med 1996; 15:827-833.[Abstract]
  17. Ramos IM, Taylor KJ, Kier R, Burns PN, Snower DP, Carter D. Tumor vascular signals in renal masses: detection with Doppler ultrasonography. Radiology 1988; 168:633-637.[Abstract/Free Full Text]
  18. Ozbek SS, Arkun R, Killi R, Memis A, Dagdeviren A, Sevinc E. Image-directed color Doppler ultrasonography in the evaluation of superficial solid tumors. J Clin Ultrasound 1995; 23:233-238.[Medline]
  19. Van deer Woude HJ, Vanderschueren G. Ultrasound in musculoskeletal tumors with emphasis on its role in tumor follow up. Radiol Clin North Am 1999; 37:753-766.[CrossRef][Medline]
  20. Declaration of Helsinki. World Med J 1982; 29:86-88.[Medline]
  21. Less JR, Skalak TC, Sevick EM, Jain RK. Microvascular architecture in a mammary carcinoma: branching patterns and vessel dimensions. Cancer Res 1991; 51:265-273.[Abstract/Free Full Text]
  22. Robbin ML, Oser RF, Allon M, et al. Hemodialysis access graft stenosis: US detection. Radiology 1998; 208:655-661.[Abstract/Free Full Text]
  23. Gooding GA, Perez S, Rapp JH, Krupski WC. Lower-extremity vascular grafts placed for peripheral vascular disease: prospective evaluation with duplex Doppler sonography. Radiology 1991; 180:379-386.[Abstract/Free Full Text]
  24. Bandyk DF, Cato RF, Towne JB. A low flow velocity predicts failure of femoropopliteal and femorotibial bypass grafts. Surgery 1985; 98:799-809.[Medline]
  25. Bartko JJ, Carpenter WT. On the methods and theory of reliability. J Nerv Ment Dis 1976; 163:307-317.[Medline]
  26. Gross ST. The kappa coefficient of agreement for multiple observers when the number of subjects is small. Biometrics 1986; 42:883-893.[CrossRef][Medline]
  27. Cyr L, Francis K. Measures of clinical agreement for nominal and categorical data: the kappa coefficient. Comput Biol Med 1992; 22:239-246.[CrossRef][Medline]
  28. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977; 33:159-174.[CrossRef][Medline]
  29. Galen S, Gambino SR. Beyond normality: the predictive value and efficiency of medical diagnoses New York, NY: Wiley, 1975.
  30. SAS Institute. JMP statistics and graphics guide Version 3. Cary, NC: SAS Institute, 1994.
  31. Murphy JM, Berwick DM, Weinstein MC, Borus JF, Budman SH, Klerman GL. Performance of screening and diagnostic tests: application of receiver operating characteristics analysis. Arch Gen Psychiatry 1987; 44:550-555.[Abstract]
  32. Enzinger FM, Weiss SW. Soft tissue tumors 3rd ed. St Louis, Mo: Mosby, 1995.
  33. Sundaram M, McLeod RA. MR imaging of tumor and tumorlike lesions of bone and soft tissue. AJR Am J Roentgenol 1990; 155:817-824.[Abstract/Free Full Text]
  34. Shubik P. Vascularization of tumors: a review. J Cancer Res Clin Oncol 1982; 103:211-226.[CrossRef][Medline]
  35. Vaupel P, Kallinowski F, Okunieff P. Blood flow, oxygen and nutrient supply, and metabolic microenvironment of human tumors: a review. Cancer Res 1989; 49:6449-6465.[Abstract/Free Full Text]
  36. Kallinowski F, Schlenger KH, Runkel S, et al. Blood flow, metabolism, cellular microenvironment, and growth rate of human tumor xenografts. Cancer Res 1989; 49:3759-3764.[Abstract/Free Full Text]
  37. Schick S, Steiner E, Gahleitner A, et al. Differentiation of benign and malignant tumors of the parotid gland: value of pulsed Doppler and color Doppler sonography. Eur Radiol 1998; 8:1462-1467.[CrossRef][Medline]
  38. Hata K, Hata T, Kitao M. Intratumoral peak systolic velocity as a new possible predictor for detection of adnexal malignancy. Am J Obstet Gynecol 1995; 172:1496-1500.[CrossRef][Medline]
  39. Dubois J, Patriquin HB, Garel L, et al. Soft-tissue hemangiomas in infants and children: diagnosis using Doppler sonography. AJR Am J Roentgenol 1998; 171:247-252.[Abstract/Free Full Text]
  40. Gamill SL, Shipkey FH, Himmelfarb EH, Parvey LS, Rabinowitz JG. Roentgenology-pathology correlative study of neovascularity. AJR Am J Roentgenol 1976; 126:376-385.[Abstract]



This article has been cited by other articles:


Home page
 Am. J. Roentgenol.Home page
L. N. Nazarian
The Top 10 Reasons Musculoskeletal Sonography Is an Important Complementary or Alternative Technique to MRI
Am. J. Roentgenol., June 1, 2008; 190(6): 1621 - 1626.
[Abstract] [Full Text] [PDF]

Home page
 J Ultrasound MedHome page
W. Jin, G. Y. Kim, B. L. Lew, D. M. Yang, H. C. Kim, J. K. Ryu, J. S. Park, and K. N. Ryu
Sonographic Findings of an Eccrine Spiradenoma: Case Report and Literature Review
J. Ultrasound Med., May 1, 2008; 27(5): 813 - 818.
[Full Text] [PDF]

Home page
 J Ultrasound MedHome page
C.-K. Chen, H.-J. Chiou, Y.-H. Chou, C.-M. Tiu, H.-T. H. Wu, S. Ma, W. Chen, and C.-Y. Chang
Sonographic Findings in Skeletal Muscle Metastasis From Renal Cell Carcinoma
J. Ultrasound Med., October 1, 2005; 24(10): 1419 - 1423.
[Full Text] [PDF]

Home page
 J Ultrasound MedHome page
S. Kaushik, T. T. Miller, L. N. Nazarian, and W. C. Foster
Spectral Doppler Sonography of Musculoskeletal Soft Tissue Masses
J. Ultrasound Med., December 1, 2003; 22(12): 1333 - 1336.
[Abstract] [Full Text] [PDF]

Home page
 ImagingHome page
I Beggs
Ultrasound of soft tissue masses
Imaging, June 1, 2002; 14(3): 202 - 208.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
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 Bodner, G.
Right arrow Articles by Jaschke, W. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bodner, G.
Right arrow Articles by Jaschke, W. R.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE