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


     

DOI: 10.1148/radiol.2392040986
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 Lucidarme, O.
Right arrow Articles by Mattrey, R. F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lucidarme, O.
Right arrow Articles by Mattrey, R. F.
(Radiology 2006;239:730-739.)
© RSNA, 2006

Experimental Studies

Angiogenesis: Noninvasive Quantitative Assessment with Contrast-enhanced Functional US in Murine Model1

Olivier Lucidarme, MD, PhD2, Yuko Kono, MD, Jacqueline Corbeil, BS, Sang-Hee Choi, MD3, Jean-Louis Golmard, MD, PhD, Judith Varner, PhD and Robert F. Mattrey, MD

1 From the Departments of Radiology (O.L., Y.K., J.C., S.H.C., R.F.M.) and Medicine (J.V.), University of California, San Diego, Calif; and Department of Biostatistics, University Pierre et Marie Curie, Paris, France (J.L.G.). From the 2002 RSNA Annual Meeting. Received June 3, 2004; revision requested August 10; revision received April 21, 2005; accepted June 17; final version accepted July 29. Supported in part by Société Française de Radiologie. Address correspondence to R.F.M., MRI Institute, 410 Dickinson St, San Diego, CA 92103 (e-mail: rmattrey{at}ucsd.edu).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Purpose: To evaluate quantitative functional ultrasonography (US) in a murine gel model by using microbubble destruction kinetics to determine whether parametric indices provided with US could help assess angiogenesis.

Materials and Methods: Institutional Animal Subjects Committee approved experiments and procedures. In 36 normal mice, two 0.4-mL gel implants were placed subcutaneously on either side of spine. One implant contained 0.5, 1.0, or 1.5 µg human basic fibroblast growth factor (bFGF) per milliliter of gel. Functional US quantitative analysis of angiogenesis with microbubble contrast agent was performed on days 3, 6, 9, and 12; histologic data were collected. Time-intensity curve of implant was fitted to mathematic decay model to calculate fractional blood volume and fraction of blood replaced per unit of time. Microvascular density (MVD) and percentage of microvascular area (MVA) were measured after anti-CD31 staining. Spearman rank order correlation was used in analyses.

Results: bFGF-containing implants induced MVD of eight, 35, 42, and 42 vessels per square millimeter on days 3, 6, 9, and 12, respectively; in controls, MVD was four vessels/mm2 (P < .05 on days 6, 9, and 12). bFGF-containing implants induced percentage MVA of 2%, 5%, 20%, and 27%, respectively; in controls, it was 0.5% (P < .05). Maximum enhancement was significantly increased in bFGF implants (23.3 gray level ± 14.1 [standard deviation]) compared with controls (11.0 ± 5.5, P < .001). Implants containing bFGF showed poor correlations between fractional blood volume and MVD (r2 = 0.42) or percentage MVA (r2 = 0.51) at US. There was no correlation between microbubble velocity and MVD (r2 < 0.05) or percentage MVA (r2 < 0.13).

Conclusion: Functional US perfusion parameters do not correlate with current histologic indices for quantifying angiogenesis. MVD, as a histologic quantitative measurement of angiogenesis, may not be an appropriate standard for contrast-enhanced imaging that relies on perfused neovessels.

© RSNA, 2006


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Angiogenesis has become a particularly important marker for cancer biology research and tumor-directed therapy. Although many antiangiogenic therapies are undergoing evaluation in clinical trials (1–6), the methods to prove therapeutic efficacy are lacking. Because antiangiogenic therapy may not lead to substantial tumor mass or tumor volume reduction, especially soon after therapy, conventional measurements of response (7,8) may be insensitive or markedly delayed even when there may be a substantial therapeutic effect. More important, with the ability to measure the functional effects of antiangiogenic drugs, therapeutic regimen optimization would be easily and rapidly achieved. The direct assessment of angiogenesis requires tissue biopsy, which not only is susceptible to sampling errors but also is not feasible for serial monitoring of patients (9). Therefore, validation and standardization of new noninvasive monitoring techniques for antiangiogenic therapy is a major need in this field.

Functional ultrasonography (US) with a contrast medium may provide an index of angiogenesis in vivo. US is extremely sensitive to microbubbles, and 2–3-µm microbubbles are limited to the intravascular space. Although functional US can provide a measurement of relative tissue blood flow and fractional blood volume in tissue (10–14), additional experimental studies must be conducted to evaluate whether this modality can be used to provide a quantitative measurement of angiogenesis (10,15,16). There are two techniques to quantify vascular indices with microbubble US contrast media, the microbubble destruction-replenishment model (13) and the destruction kinetic model (17). Each has its own advantages and disadvantages. One of the major differences between the two techniques is the time required to acquire a sufficient data set to allow the calculation of transit time and fractional blood volume. The destruction-replenishment model requires a minimum of 15–20 seconds; however, it has been shown that the destruction kinetic model can reliably provide the fractional blood volume and microbubble velocity indices in 1–4 seconds (18).

Angiogenesis is a complex process, and even though tumors are the ultimate model for the study of angiogenesis, tumor growth is influenced by many local and systemic factors that introduce variability within the same tumor type and among different tumor types (19). Further, given the heterogeneity of tumor histologic characteristics, it is difficult to confirm that the region imaged is the same region assessed histologically. For these reasons, a gel model (Matrigel; Becton Dickinson Labware, Bedford, Mass) has been used extensively in molecular biology to study the biochemical, molecular, and cellular events associated with angiogenesis in vitro and in vivo (20–22). The gel is acellular, is made from the basement membrane of mouse tumors, and is nearly completely depleted of growth factors. It is a complex mixture of basement membrane proteins that include laminin, type IV collagen, entactin, and proteoheparan sulfate (22). The material is a liquid at 4°C, but it forms a gel at a slightly higher temperature. The gel makes an ideal model for US because it has been shown that 0.4 mL of it produces a hypoechoic space that remains unchanged during a 12-day observation period. In addition, when basic fibroblast growth factor (bFGF) is added, an angiogenic plug forms around and penetrates the gel implant (22,23). This plug lends itself to US before and after contrast agent administration, providing an ideal noncomplicated anechoic and acellular background with which to correlate angiogenesis and functional US (23).

The purpose of our current study was to evaluate quantitative functional US for assessment of angiogenesis in a murine gel model by using microbubble destruction kinetics and to determine whether parametric indices provided with US could be used to assess angiogenesis.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 References
 
The Philippe Foundation, Paris, France, and IMCOR Pharmaceutical, San Diego, Calif, provided the contrast agent. The authors had control of the experimental design, the data, and the information submitted for publication.

Animal Model
We used a murine gel angiogenesis model adapted for US (22,23). Growth factor–depleted gel was cooled to a liquid consistency at 4°C. Then 0.4 mL was withdrawn into a cooled needle and syringe and injected into the mouse in the subcutaneous space adjacent to the spine as a control implant. At the same time, 0.4 mL of cooled growth factor–depleted gel mixed with a given concentration of human bFGF (R&D Systems, Minneapolis, Minn) was injected on the opposite side of the spine. The injection was performed at the same vertebral level near the lumbar spine as was selected for the control implant, and it was below the rib cage so that the kidneys could be included on the same transverse image as was obtained of the gel implant.

We examined 36 6–7-week-old National Institutes of Health Swiss mice (Harlan Sprague Dawley, Indianapolis, Ind) that weight 20–25 g. To obtain a wide range of angiogenic response, a technician impregnated the bFGF-containing gel implants with three concentrations of human bFGF: 0.5 (n = 12), 1.0 (n = 12), or 1.5 (n = 12) micrograms per milliliter of gel that corresponded to a volume of 25, 50, and 75 µL of bFGF injected per 0.4 mL of gel. For each mouse, the side and the amount of bFGF were randomly chosen. To obtain additional variability of angiogenic response, three mice randomly chosen from each group with gel containing specified concentrations of bFGF were imaged and then sacrificed on days 3, 6, 9, and 12 after gel injection. Randomization was performed by the same technician who prepared the gel solutions and who was involved in no other part of the study. Without specific criteria, he assigned each mouse to a group to obtain exactly 12 mice in each concentration group, and he assigned three mice from each group to be sacrificed on days 3, 6, 9, and 12.

In accordance with the U.S. Department of Agriculture, Department of Health and Human Services, and National Institutes of Health policies regarding the humane care and use of laboratory animals, our Institutional Animal Care and Use Committee reviewed and approved all experiments and procedures.

Imaging Protocol
Two investigators (O.L., J.C.) blinded in regard to the side of the control implant and in regard to the bFGF concentration added in the implants performed all the imaging procedures together. Mice were anesthetized with intraperitoneal injections of ketamine (Ketaset; Fort Dodge Animal Health, Fort Dodge, Iowa) (50 mg per kilogram body weight) and acepromazine (Xylazine 100 Injection; Butler, Columbus, Ohio) (1 mg/kg) and placed in the prone position in a box made of expanded rigid polystyrene plastic with the head and tail exposed. The box was then filled with US gel. The gel was heated to 36°C and contained 0.5 g of cellulose per 100 mL of gel (24). A clinical US unit (Sonoline Elegra; Siemens Medical Solutions, Issaquah, Wash) equipped with a 13-MHz linear transducer (VFX 13–5; Siemens Medical Solutions) was used. The mouse was imaged through a 2-cm gel standoff in the transverse plane. The transducer was mechanically stabilized in a position that allowed good visualization of both gel implants. We used intermittent phase-inversion imaging and US contrast agent–dedicated software (Ensemble Contrast Imaging; Siemens Medical Solutions) with three protocol settings, setting 1, setting 2, and setting 3, that had different bubble-clearing capability (Table 1). The setting 1 technique used a high-transmit frequency that is less destructive than the lower frequency (24,25) and a slower frame rate, which allowed more microbubbles to reenter the region between frames. The imaging parameters for settings 2 and 3 were identical, except for a greater transmit power and lower receiver gain with setting 3, features that made it the most destructive setting. Time gain compensation settings were fixed in the same position for settings 1–3 during the entire study.


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

  		Table 1. US Settings

 
For each of the three imaging protocols, a bolus of 0.1 mL of a microbubble US contrast agent (Imagent; IMCOR Pharmaceutical), followed by a 0.1-mL saline flush, was manually injected for 1–2 seconds by using a 1-mL syringe that was connected to a 27-gauge needle, and the needle was inserted into the tail vein (O.L., J.C.). This microbubble US contrast agent encapsulates perfluorohexane vapor in a thin lipid membrane and contains (5 to 8) x 108 microbubbles per milliliter of solution (26). The 0.1-mL dose is 350 times larger than that for human use. It was chosen to obtain a visually assessable enhancement. To control the quality of the injections, we observed the enhancement of the cortex of both kidneys, which were always included in the imaging plane. We also visually monitored the opacity of the liquid during the injection to prevent microbubble destruction in the syringe. If microbubbles are destroyed, the suspension becomes transparent and the contrast agent loses its effect.

Three frames at each of the three imaging settings were acquired at baseline prior to any contrast agent injection. After a 30-second delay after contrast agent administration, 20 frames were acquired intermittently at fixed intervals as prescribed in each imaging protocol. After a 5-minute delay between injections, to allow for microbubble clearance before the second and third injections, 10 frames were acquired at maximum power, lowest frequency, and highest frame rate to clear all potentially remaining microbubbles from the implants. The delay of 30 seconds after injection was chosen to allow complete vascular filling by microbubbles.

For all imaging protocols, we used a linear postprocessing curve (named "display curve 4" in the US machine used) that led to a linear relationship between the logarithm of echo amplitude (in decibels) and the gray-scale values. We also used zero persistence that eliminates averaging with prior frames. Digital images were transferred to a computer (Optiplex GX270; Dell, Round Rock, Tex) for postprocessing.

Imaging Data and Kinetic Model
With the use of a morphometric analysis program (Scion Image; Scion, Frederick, Mass), a region of interest (ROI) was drawn by one of the authors (O.L.) along the perimeter of each of the two gel implants on the nonenhanced images, and the ROI was automatically positioned over all implants on subsequent images, with minor adjustments to correct for respiratory motion. Mean video intensity of each ROI on each image was recorded. Three observers, with 2–6 years of experience with US microbubble imaging (S.H.C., O.L., Y.K.), who were blinded in regard to the sides on which the control implants and the bFGF-containing implants were placed and the concentration, reviewed the cine loop of each acquisition and consensually determined the presence of focal enhancement within each gel implant. An ROI was drawn around the focus of enhancement when one was present, and the mean video intensity ± standard deviation of the ROI was measured on all images that were obtained with the same imaging protocol. Contrast enhancement was defined as the difference between the mean video intensity after injection and the average baseline value.

The time-intensity curve for each imaging protocol was plotted, and a least square fit was calculated by using a kinetic model that was described (17) and validated previously (18). Briefly, the model was built on the basis of three assumptions: The number of microbubbles entering the region is constant during the imaging period, which in this experiment was either 4 or 8 seconds; the fraction of microbubbles cleared with each frame is constant; and no microbubble entrapment occurs within the ROI. At equilibrium, when the vascular space is completely filled and the number of microbubbles entering is equal to the number exiting the region, enhancement can be used as an indicator for fractional blood volume. In this experiment, conditions for this last assumption were obtained on the first frame after the 30-second delay after contrast agent administration when the enhancement was maximum. With each subsequent frame, the enhancement decays until it reaches a new equilibrium when the number of microbubbles entering the field is equal to the sum of the microbubbles being cleared with each frame and those flowing out of the field. When the microbubble-clearing capability of an imaging technique is either too low—the time-intensity curve remains flat—or too high—the time-intensity curve decays to baseline in 1 or 2 frames—the model cannot fit the observed data (18). When the model fits the observed time- intensity curve, it yields two parameters: the fraction of microbubbles cleared with each frame that is reported as the fraction of microbubbles cleared per unit time ({lambda}) and the fraction of vascular volume replaced per second, which is also the inverse of microbubble transit time in the plane imaged (1/{tau}) or mean microbubble velocity. In this experiment, we considered that the kinetic model adequately described the observed decay curve when the goodness of the fit (coefficient of determination rd2) was 0.80 or greater.

Histologic Quantification of Angiogenesis
Immediately after imaging, mice were euthanized, and the gel implants were extracted and photographed. In consensus, two observers who were blinded to the bFGF concentration and the results of US (S.H.C., Y.K.) assigned scores for the angiogenesis response macroscopically as follows: score 1, the implant was not penetrated by vessels; score 2, implants were peripherally penetrated by vessels; or score 3, vessels penetrated deep into the implant. Axial frozen sections of each implant that were prepared in a manner to approximate the imaging plane were obtained and stained with anti-CD31. With the expertise of one of the authors (J.V.) who had more than 10 years of experience with microvascular density (MVD) counts, particularly in gel implant plugs, two observers blinded to the bFGF concentration and the results of US (O.L., J.C.) independently counted the MVD and the fraction of the implant area involved with vessels, which we called percentage of microvascular area (MVA), by using light microscopy. The MVD score, which reflects local proliferation of neovessels, was assessed in vascular hot spots according to Vermeulen et al (27) and Weidner et al (28,29). Hot spots correspond to areas that show stronger CD31 staining and, consequently, a higher vascular density than the rest of the tissue studied. Hot spots were located by scanning the gel implant sections with a x40 magnification. Once identified, five fields with a x200 magnification were randomly chosen within each hot spot, and each endothelial cell or cell cluster that showed antibody staining and that was clearly separated from adjacent clusters was counted. We calculated the MVD (vessels per square millimeter) for each gel implant by dividing the total count from the five hot spots by the area of the five fields with a x200 magnification (5 x 0.74 mm2), and then we averaged the results obtained by the two observers. The percentage of MVA score was assessed semiquantitatively by superimposing a grid of 1 x 1 mm2 at low magnification (x4). The ratio of the area containing neovessels stained with CD31 and the area of the implant was calculated. Percentage of MVA was the average of the ratio of the results obtained by the two observers.

Statistical Analysis
Results are expressed as a percentage or as the mean ± standard error of the mean, except for the video intensity measured in the ROI drawn around a focus of enhancement in the gel implant, which was expressed as the mean ± standard deviation. Correlations between variables were determined by using the Spearman rank order coefficient of correlation (r). Interobserver agreements were assessed by using the intraclass correlation coefficient of agreement (ri) (30,31). Although there are no universal standards, ri values of less than 0.40 are considered low, and ri values of greater than 0.75 are considered high (32). A paired Student t test was used to compare the control implants and bFGF-containing implants within mice, and an unpaired Student t test was used for comparisons among groups. Because MVD, percentage of MVA, and maximum enhancement have irregular distribution, the effect of concentration and time on these variables was tested after dichotomization of each of the three criteria by using the median of each of the three criteria as the threshold value. The relationships between time, concentration, and each of the three binary variables were then tested by using logistic models. All P values were two tailed and were considered to indicate a significant difference if they were less than .05. Calculations were performed with software (Statistica, version 5.5, 1999, StatSoft, Tulsa, Okla; SAS, version 8, 2000, SAS Institute, Cary, NC).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 References
 
Five mice were eliminated from the analysis because of contrast agent injection failure (n = 1), a flat gel implant (n = 2), or migration (n = 2), which prevented either adequate separation or viewing of the two implants on the same image. Results were based on the 31 mice that completed the protocol, with 11, nine, and 11 mice having received 0.5, 1.0, and 1.5 µg/mL of bFGF, respectively, and seven, eight, nine, and seven that were examined on days 3, 6, 9, and 12, respectively, after gel implantation (Table 2).


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

  		Table 2. Number of Mice Examined as a Function of bFGF Concentration per Milliliter of Gel and Delay after Implantation

 
Angiogenic Response
None of the 31 control implants had an appreciable vascular response. Of the 31 implants containing bFGF, seven (23%) had no angiogenic response at gross inspection, and these findings were confirmed histologically but were still considered when we generated MVD and percentage of MVA scores; 13 (42%) had a vascular response that remained at the periphery of the implant; and 11 (35%) had vessels that penetrated deep within the implant. Four of seven implants without an angiogenic response were among those containing 0.5 µg of bFGF per milliliter. The remaining three were among those containing 1.0 µg of bFGF per milliliter; two were examined on day 3 and one was examined on day 12. Implants in 31 animals, therefore, provided a wide range of angiogenic responses for analysis.

MVD (32.4 ± 28.8 vs 3.7 ± 5.1) in bFGF-containing implants compared with control implants and percentage of MVA scores (13.6% ± 15.2 vs 0.5% ± 1.2) in these two groups were significantly higher (P < .001). For all bFGF concentrations, MVD and percentage of MVA scores were significantly higher when time following gel implantation increased (P = .02 and .006, respectively) (Fig 1). For all times after gel implantation, a trend toward increased MVD and percentage of MVA with increasing concentration was also seen (Fig 1), but significance was not reached unless the 3-day time was eliminated (P = .04).


Figure 1
View larger version (20K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1a: Graphs show (a) MVD and (b) percentage of MVA scores ± standard error of the mean (SEM) (error bars) as a function of days following gel implantation and bFGF concentration. For all bFGF concentrations, MVD and percentage of MVA scores significantly increased as a function of days following gel implantation.

 

Figure 1
View larger version (19K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1b: Graphs show (a) MVD and (b) percentage of MVA scores ± standard error of the mean (SEM) (error bars) as a function of days following gel implantation and bFGF concentration. For all bFGF concentrations, MVD and percentage of MVA scores significantly increased as a function of days following gel implantation.

 
The intraclass correlation coefficients for the two observers were moderate (ri = 0.72) for the MVD index and high (ri = 0.88) for the percentage of MVA measurement.

US Results
Qualitative assessment.—At baseline, the 31 control implants and the 20 bFGF-containing implants with no or only peripheral vessel involvement were anechoic. Among the 11 bFGF-containing implants with vascular penetration deep into the implant at gross examination, nine had a mildly echogenic focus within the implant visible at baseline that corresponded to the angiogenic focus at macroscopic examination, and two were anechoic. After contrast agent administration, the cortex of both kidneys exhibited strong enhancement in all cases, and diffuse peripheral enhancement occurred with all implants but was more intense for implants containing bFGF (Fig 2). In addition, the nine implants with a mildly echogenic focus demonstrated internal enhancement. Figure 2 is a typical example of an implant with deep vascular involvement.


Figure 2
View larger version (191K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a: (a) Transverse functional US image of back of mouse in prone position shows gel implants (*) placed 9 days earlier on both sides of the spine (S). Left implant contained 1.0 µg/mL bFGF. Intense enhancement of both kidney (K) cortices served as visual control of quality of contrast agent injection. Top: Before imaging, bFGF implant (arrows) had an echogenic focus. Bottom: In 1st frame acquired 30 seconds after contrast agent administration with setting 2, bFGF implant (arrows) enhanced intensely. (b) Graph shows corresponding video intensity curves when the ROI was drawn around the entire implant. Maximum enhancement was 42 gray-scale level for the left and 9 gray-scale level for the right implant. The least square fit of the kinetic model to the bFGF implant data had an rd2 of 0.91. Coefficient of destruction ({lambda}) and microbubble velocity (1/{tau}) were 0.10 and 0.005, respectively. The kinetic model could not fit the data of the control implant (rd2 = 0.61), and no parameters could be calculated. IV = intravenous injection of contrast agent. (c) Corresponding gross specimen (1, 2) and specimen at x5 magnification (3, 4) of CD31-stained section of control (1, 3) and bFGF-containing (2, 4) gel implants. MVD and percentage of MVA were 0 and 0% for the control implant and 62.2 vessels per square millimeter and 22% for the bFGF-containing implant, respectively.

 

Figure 2
View larger version (18K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b: (a) Transverse functional US image of back of mouse in prone position shows gel implants (*) placed 9 days earlier on both sides of the spine (S). Left implant contained 1.0 µg/mL bFGF. Intense enhancement of both kidney (K) cortices served as visual control of quality of contrast agent injection. Top: Before imaging, bFGF implant (arrows) had an echogenic focus. Bottom: In 1st frame acquired 30 seconds after contrast agent administration with setting 2, bFGF implant (arrows) enhanced intensely. (b) Graph shows corresponding video intensity curves when the ROI was drawn around the entire implant. Maximum enhancement was 42 gray-scale level for the left and 9 gray-scale level for the right implant. The least square fit of the kinetic model to the bFGF implant data had an rd2 of 0.91. Coefficient of destruction ({lambda}) and microbubble velocity (1/{tau}) were 0.10 and 0.005, respectively. The kinetic model could not fit the data of the control implant (rd2 = 0.61), and no parameters could be calculated. IV = intravenous injection of contrast agent. (c) Corresponding gross specimen (1, 2) and specimen at x5 magnification (3, 4) of CD31-stained section of control (1, 3) and bFGF-containing (2, 4) gel implants. MVD and percentage of MVA were 0 and 0% for the control implant and 62.2 vessels per square millimeter and 22% for the bFGF-containing implant, respectively.

 

Figure 2
View larger version (138K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2c: (a) Transverse functional US image of back of mouse in prone position shows gel implants (*) placed 9 days earlier on both sides of the spine (S). Left implant contained 1.0 µg/mL bFGF. Intense enhancement of both kidney (K) cortices served as visual control of quality of contrast agent injection. Top: Before imaging, bFGF implant (arrows) had an echogenic focus. Bottom: In 1st frame acquired 30 seconds after contrast agent administration with setting 2, bFGF implant (arrows) enhanced intensely. (b) Graph shows corresponding video intensity curves when the ROI was drawn around the entire implant. Maximum enhancement was 42 gray-scale level for the left and 9 gray-scale level for the right implant. The least square fit of the kinetic model to the bFGF implant data had an rd2 of 0.91. Coefficient of destruction ({lambda}) and microbubble velocity (1/{tau}) were 0.10 and 0.005, respectively. The kinetic model could not fit the data of the control implant (rd2 = 0.61), and no parameters could be calculated. IV = intravenous injection of contrast agent. (c) Corresponding gross specimen (1, 2) and specimen at x5 magnification (3, 4) of CD31-stained section of control (1, 3) and bFGF-containing (2, 4) gel implants. MVD and percentage of MVA were 0 and 0% for the control implant and 62.2 vessels per square millimeter and 22% for the bFGF-containing implant, respectively.

 
Quantitative assessment.—Regardless of bFGF concentration or time after implantation, the enhancement, which was used as an indicator for fractional blood volume, measured on the first frame after the 30-second delay after contrast agent administration was significantly greater in implants containing bFGF than in their respective control implants with all imaging settings: setting 1 (the least microbubble-destructive technique), 9.2 ± 1.8 versus 4.6 ± 0.7 (P = .01); setting 2 (the technique with intermediate microbubble destruction), 23.3 ± 2.5 versus 11.0 ± 1.0 (P = .001); and setting 3 (the most microbubble-destructive technique of the three), 23.4 ± 2.5 versus 13.9 ± 1.4 (P = .001).

There was a significant trend toward increased maximum enhancement with the increased number of days after implantation (P = .006); however, these differed with the imaging technique used (Fig 3). With setting 2, a difference was observed between the bFGF-containing implant and its respective control implant for days 6, 9, and 12. With setting 3, a difference was observed for days 9 and 12, and with setting 1, a difference was observed on only day 12. A trend toward increased maximum enhancement with increased concentration did not reach the significant level.


Figure 3
View larger version (32K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3: Graph shows mean maximum enhancement (gray-scale level) ± standard error of the mean (SEM) (error bars) in all control and bFGF-containing implants within days following gel implantation for settings 1–3 (S1–S3) (Table 1). Significant differences in enhancement were observed as a function of days following implantation. * = P < .05 and ** = P < .01, relative to control implant.

 
When we attempted to show a relationship between maximum enhancement of the entire area of each bFGF-containing implant and MVD or that between maximum enhancement and percentage of MVA, poor correlations were observed: for setting 1, r2 = 0.19 (P < .01) and r2 = 0.17 (P < .001); for setting 2, r2 = 0.42 (P < .001) and r2 = 0.51 (P < .001) (Fig 4); and for setting 3, r2 = 0.28 (P < .002) and r2 = 0.37 (P < .001), respectively. When we attempted to show a relationship between maximum enhancement of only the focus of internal enhancement that was observed in nine of 11 implants with deep vascular penetration and MVD, no correlation was observed with any of the imaging techniques. For example, r2 was less than 0.001 (P = .94) with setting 2 (Fig 5).


Figure 4
View larger version (19K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4a: Graphs show correlation between maximum enhancement of entire implant obtained with setting 2 and (a) MVD and (b) percentage of MVA. Poor positive correlations were observed, slightly greater for percentage of MVA than for MVD.

 

Figure 4
View larger version (20K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4b: Graphs show correlation between maximum enhancement of entire implant obtained with setting 2 and (a) MVD and (b) percentage of MVA. Poor positive correlations were observed, slightly greater for percentage of MVA than for MVD.

 

Figure 5
View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5: Correlation between mean maximum enhancement obtained ± standard deviation (SD) (error bars) measured in the nine ROIs drawn around the internal focus of enhancement seen in the bFGF-containing gel implants with setting 2 and highest MVD.

 
Kinetic Model Results
The parameters given by the kinetic model were considered only when the model was able to fit the time-intensity curve with the least square fit, yielding a coefficient of determination of 0.80 or greater. The model was able to fit the observed time-intensity curves in two (6%) of 31 bFGF-containing implants imaged with setting 1; 21 (68%) of 31, with setting 2; and 28 (90%) of 31, with setting 3. The {lambda} (clearing coefficient) and 1/{tau} (fraction of blood replaced per unit of time) were then analyzed for settings 2 and 3.

The microbubble-clearing coefficient that is strictly dependent on transmit parameters for the microbubble system used was four times greater for the more destructive setting 3 than it was for setting 2 (0.66 ± 0.46 vs 0.16 ± 0.08, P < .001). The cleared fraction of microbubbles with each frame was correlated with neither MVD nor percentage of MVA. All r2 values were less than 0.18 (P > .05).

The fraction of blood replaced per unit of time, a physiologic parameter that is independent of imaging technique, was similar for both settings 2 and 3 (mean, 0.03 ± 0.03 vs 0.04 ± 0.005) and did not correlate with either MVD or percentage of MVA. All r2 values were less than 0.13 (P > .05).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
References
 
Blood flow, blood volume, vascular density, and vascular permeability are parameters functionally and anatomically associated with tumor angiogenesis (33). Microbubbles have a true blood pool distribution, which gives functional US the capability to quantify local blood flow and fractional blood volume (12,13,17) but not vascular permeability. In this study, we aimed to assess the potential role of functional US in the quantification of angiogenesis.

Biocompatible polymer matrix impregnated with growth factors and implanted subcutaneously in mice, rats, and pigs, as well as the chorioallantoic membrane of chick embryos, has been used extensively to provide a stable and reproducible system to study the different biochemical and cellular events associated with angiogenesis (34). Properly prepared gel implants are homogeneous and acellular and appear anechoic at US, and they provide an ideal imaging and histologic background for evaluation of the enhancement of angiogenic foci induced by microbubbles. When the gel was impregnated with bFGF, it produced an angiogenic response that increased over time. This induced angiogenic process involved only the periphery in 42% (13 of 31) of implants that remained anechoic and penetrated deep into the implants in 35% (11 of 31), appearing as a faintly echogenic focus in nine of 11 implants. An imaging plane distant from the angiogenic focus was likely chosen in the two other mice that had deep angiogenic foci. These foci must have been too faint to be observed before the contrast agent was administered when the transducer was mechanically fixed over the implants.

Because of the very high heart rate, small blood volume, and large microbubble load per kilogram (4 x 109 microbubbles per kilogram), we considered that equilibrium was reached 30 seconds after injection. When we assumed that no microbubbles were trapped when equilibrium was reached, the number of microbubbles entering the region equaled the number exiting the region. That no entrapment occurs in tissue microvessels has been demonstrated in other studies (13,14) with the destruction-reperfusion model in which a plateau of maximal enhancement is reached with delay times of 10–15 seconds or longer. Therefore, the degree of enhancement observed on the first frame 30 seconds after injection of contrast agent theoretically reflects the fractional blood volume present in the region. We assumed that blood microbubble concentration was similar among all animals because all animals were essentially of the same age and weight and all received the same contrast agent dose. Therefore, differences in enhancement observed on the first frame obtained with a given imaging technique among animals reflect differences in fractional blood volume within implants.

The three concentrations of bFGF and the four times for imaging provided a widespread angiogenic response that allowed the correlation of the functional US parameters with MVD and percentage of MVA within a wide range of values. The degree of enhancement observed on the first image and the fraction of blood replaced per second (1/{tau}) given by the kinetic model did not correlate with percentage of MVA or MVD, which had been demonstrated to be a significant and independent prognostic indicator in many cancers and included those of the breast, prostate, ovary, stomach, and colon and melanoma (35–39). Three hypotheses can be put forward to explain this lack of correlation.

First, imprecise correlation between the imaging plane and the histologic section could have occurred.

Second, although it is expected that intense microvascular proliferation will lead to a greater number of vessels and therefore greater fractional blood volume and greater blood flow, these indices may not be totally correlated and are likely out of phase during rapid proliferation or antiangiogenic therapy. MVD, assessed with anti-CD31 staining, is a measurement of all capillaries, which include those that are occluded or not yet connected to the systemic circulation (27). These vessels do not induce any change in local blood volume and blood flow, whereas US depicts only perfused regions. This phenomenon may be more of a limiting factor in this artificial nontumor model, where vessels grow toward the center of the avascular implant rather than along the perfused tumor periphery. In addition, in this study, only one growth factor was added. Hence, the number of immature nonperfused vessels could be higher in the gel implant than it is in a real tumor. This hypothesis is supported by findings of Forsberg et al (15), who reported that no correlation between enhanced color Doppler density (number of colored pixels per unit of surface) and the extent of angiogenesis assessed with CD31 markers was observed in 14 mice. Similarly, Iordanescu et al (40) did not find any correlation between MVD (assessed with CD31 antibodies) and vascularity seen at contrast material–enhanced color Doppler US. Kim et al (41) also reported that a poor linear correlation was found between enhanced power Doppler US and MVD assessed with CD34 (r2 = 0.30, P = .003) in 29 breast carcinomas. Huber et al (42) observed no significant correlation between color-enhanced Doppler density and MVD assessed with CD31 in 40 fibroadenomas. Findings in only one study (43) indicated that there was a good correlation between contrast-enhanced transrectal color Doppler intensity and MVD (assessed with anti–factor VIII) in prostate cancer specimens (r2 = 0.977, P < .001).

Third, the gel implant does not need any oxygen or nutrients; thus, stimulating factors for increasing the blood flow may be lacking, leading to a constant volume of blood replaced per unit of time (1/{tau}) despite the development of neovessels.

These data strongly suggest that MVD as a measurement of angiogenesis is not an ideal standard with which to compare perfusion parameters that result from contrast agent–based time-intensity data. This does not imply, however, that such data are or will be of no utility for assessing tumor perfusion or alteration in relation to tumor perfusion in response to therapy.

Better correlations with US enhancement would be expected if vessels are stained with lectin that is injected before euthanasia or if red blood cells are labeled or counted or hemoglobin concentration is measured to quantify the functional capillaries (44).

The kinetic model used in this study and validated in vitro (18,45) is based on the rate of decay of the observed enhancement that can be evaluated within 1–5 seconds, depending on the frame rate and transmit power used. This model, however, requires that a decay in enhancement occurs over several frames, and, therefore, the appropriate transmit power and frame rate must be chosen. Insufficient decay in enhancement occurred with setting 1, the least destructive setting, and this occurrence led to limitation of the ability to fit the data to data in only two of 31 implants. The most destructive setting, setting 3, allowed the model to fit 28 of the 31 decay curves. It is possible that setting 2 imaging parameters yielded better statistics on the first frame than did setting 3 parameters because of the lesser near–acoustic field noise that accompanies high transmit power.

There were two limitations in this study: First, correlations with MVD could be underestimated since the two observers who assessed the MVD were not pathologists and had limited experience in performing MVD counts. Several authors have underlined the high interobserver variability in MVD assessment even with experienced pathologists; however, counting of vessels for assessment of MVD in gel implants is easier than it is in biologic tissue because gel is totally translucent without any cells in the background. Thus, CD31 staining is easy to detect, with no false-positive results. In our study, the respective values for interobserver agreement for MVD and percentage of MVA assessment were moderate and high. Second, we measured the enhancement of the entire implant, whereas only a peripheral and/or small area of the implant was affected by angiogenesis, and MVD is the number of capillaries per square area of highest proliferation. This factor may have caused us to overlook different stages of angiogenesis. When we limited the analysis to the nine implants with deep vascular penetration and correlated the enhancement of each focus to the MVD of each focus, however, we also observed no correlation.

In conclusion, our study results indicate that, even when an appropriate correlation of time-intensity curves from functional US with the kinetic model chosen exists, the perfusion parameters yielded by this model do not correlate with the current histologic standard used to clinically quantify angiogenesis. Although this observation could reflect the limitations of our model and experimental design, it also underscores the complexity of angiogenesis and the limitations of correlating functional imaging parameters with morphologic histologic indices.

Practical application: Peak enhancement on contrast-enhanced functional US images, when appropriate US protocol settings are chosen, correlates poorly with the percentage of cross-sectional area involved with neovessels in a given ROI (percentage of MVA) and the MVD assessed in the hot spot on a histologic slice. These histologic indices currently used in clinical practice may not reflect the functional microcirculation and may not be the appropriate standards that contrast-enhanced imaging should be judged against.


   FOOTNOTES
 

Abbreviations: bFGF = basic fibroblast growth factor • MVA = microvascular area • MVD = microvascular density • ROI = region of interest

2 Current address: Laboratoire d'Imagerie Parametrique, UMR 7623 CNRS-University Paris VI and Assistance Publique Hôpitaux de Paris, France. Back

3 Current address: Department of Radiology and Center for Imaging Science, Sungkyunkwan University, Seoul, Korea. Back

See Materials and Methods for pertinent disclosures.

Author contributions: Guarantors of integrity of entire study, O.L., Y.K., R.F.M.; 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, O.L., J.V., R.F.M.; experimental studies, O.L., Y.K., J.C., S.H.C., J.V., R.F.M.; statistical analysis, J.L.G.; and manuscript editing, O.L., Y.K., J.C., J.L.G., J.V., R.F.M.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 

  1. Figg WD, Dahut W, Duray P, et al. A randomized phase II trial of thalidomide, an angiogenesis inhibitor, in patients with androgen-independent prostate cancer. Clin Cancer Res 2001;7:1888–1893.[Abstract/Free Full Text]
  2. Herbst RS, Langer CJ. Epidermal growth factor receptors as a target for cancer treatment: the emerging role of IMC-C225 in the treatment of lung and head and neck cancers. Semin Oncol 2002;29:27–36.[Medline]
  3. Kohn EC, Reed E, Sarosy GA, et al. A phase I trial of carboxyamido-triazole and paclitaxel for relapsed solid tumors: potential efficacy of the combination and demonstration of pharmacokinetic interaction. Clin Cancer Res 2001;7:1600–1609.[Abstract/Free Full Text]
  4. Kudelka AP, Verschraegen CF, Loyer E. Complete remission of metastatic cervical cancer with the angiogenesis inhibitor TNP-470. N Engl J Med 1998;338:991–992.[Free Full Text]
  5. Offodile R, Walton T, Lee M, Stiles A, Nguyen M. Regression of metastatic breast cancer in a patient treated with the anti-angiogenic drug TNP-470. Tumori 1999;85:51–53.[Medline]
  6. Vine AK, Brewer GJ. Tetrathiomolybdate as an antiangiogenesis therapy for subfoveal choroidal neovascularization secondary to age-related macular degeneration. Trans Am Ophthalmol Soc 2002;100:73–76.[Medline]
  7. Miller AB, Hoogstraten B, Staquet M, Winkler A. Reporting results of cancer treatment. Cancer 1981;47:207–214.[CrossRef][Medline]
  8. Therasse P, Arbuck SG, Eisenhauer EA, et al. New guidelines to evaluate the response to treatment in solid tumors: European Organization for Research and Treatment of Cancer, National Cancer Institute of the United States, National Cancer Institute of Canada. J Natl Cancer Inst 2000;92:205–216.[Abstract/Free Full Text]
  9. Schor AM, Pendleton N, Pazouki S, et al. Assessment of vascularity in histological sections: effects of methodology and value as an index of angiogenesis in breast tumours. Histochem J 1998;30:849–856.[CrossRef][Medline]
  10. Pollard RE, Sadlowski AR, Bloch SH, et al. Contrast-assisted destruction-replenishment ultrasound for the assessment of tumor microvasculature in a rat model. Technol Cancer Res Treat 2002;1:459–470.[Medline]
  11. Schlosser T, Pohl C, Veltmann C, et al. Feasibility of the flash-replenishment concept in renal tissue: which parameters affect the assessment of the contrast replenishment? Ultrasound Med Biol 2001;27:937–944.
  12. Seidel G, Claassen L, Meyer K, Vidal-Langwasser M. Evaluation of blood flow in the cerebral microcirculation: analysis of the refill kinetics during ultrasound contrast agent infusion. Ultrasound Med Biol 2001;27:1059–1064.[CrossRef][Medline]
  13. Wei K, Jayaweera AR, Firoozan S, Linka A, Skyba DM, Kaul S. Quantification of myocardial blood flow with ultrasound-induced destruction of microbubbles administered as a constant venous infusion. Circulation 1998;97:473–483.[Abstract/Free Full Text]
  14. Wei K, Le E, Bin JP, Coggins M, Thorpe J, Kaul S. Quantification of renal blood flow with contrast-enhanced ultrasound. J Am Coll Cardiol 2001;37:1135–1140.[Abstract/Free Full Text]
  15. Forsberg F, Dicker AP, Thakur ML, et al. Comparing contrast-enhanced ultrasound to immunohistochemical markers of angiogenesis in a human melanoma xenograft model: preliminary results. Ultrasound Med Biol 2002;28:445–451.[CrossRef][Medline]
  16. Schroeder RJ, Hauff P, Bartels T, et al. Tumor vascularization in experimental melanomas: correlation between unenhanced and contrast enhanced power Doppler imaging and histological grading. Ultrasound Med Biol 2001;27:761–771.[CrossRef][Medline]
  17. Wilkening W, Postert T, Federlein J, Kono Y, Mattrey R, Ermert H. Ultrasonic assessment of perfusion conditions in the brain and in the liver. In: Schneider SC, Levy M, McAvoy BR, eds. Ultrasonics Symposium Proceedings. Vol 2. New York, NY: Institute of Electrical and Electronic Engineers, 2000; 1544–1548.
  18. Lucidarme O, Kono Y, Corbeil J, Choi S, Mattrey R. Validation of ultrasound contrast destruction imaging for flow quantification. Ultrasound Med Biol 2003;29:1697–1704.[CrossRef][Medline]
  19. Jain RK. Determinants of tumor blood flow: a review. Cancer Res 1988;48:2641–2658.[Abstract/Free Full Text]
  20. Baatout S, Cheta N. Matrigel: a useful tool to study endothelial differentiation. Rom J Intern Med 1996;34:263–269.[Medline]
  21. Kleinman HK, McGarvey ML, Hassell JR, et al. Basement membrane complexes with biological activity. Biochemistry 1986;25:312–318.[CrossRef][Medline]
  22. Passaniti A, Taylor RM, Pili R, et al. A simple, quantitative method for assessing angiogenesis and antiangiogenic agents using reconstituted basement membrane, heparin, and fibroblast growth factor. Lab Invest 1992;67:519–528.[Medline]
  23. Lucidarme O, Nguyen T, Kono Y, et al. Angiogenesis model for ultrasound contrast research: exploratory study. Acad Radiol 2004;11:4–12.[CrossRef][Medline]
  24. Frinking PJ, Céspedes E, Kirkhorn J, Torp H, de Jong N. A new ultrasound contrast imaging approach based on the combination of multiple imaging pulses and a separate release burst. IEEE Trans Ultrason Ferroelectr Freq Control 2001;48:643–651.[CrossRef][Medline]
  25. Walker KW, Pantely GA, Sahn DJ. Ultrasound-mediated destruction of contrast agents: effect of ultrasound intensity, exposure, and frequency. Invest Radiol 1997;32:728–734.[CrossRef][Medline]
  26. Perflexane (AF0150, AFO 150, Imagent, Imavist). Drugs R D 2002;3:306–309.[CrossRef][Medline]
  27. Vermeulen PB, Gasparini G, Fox SB, et al. Quantification of angiogenesis in solid human tumours: an international consensus on the methodology and criteria of evaluation. Eur J Cancer 1996;32A:2474–2484.[CrossRef]
  28. Weidner N, Folkman J, Pozza F, et al. Tumor angiogenesis: a new significant and independent prognostic indicator in early-stage breast carcinoma. J Natl Cancer Inst 1992;84:1875–1887.[Abstract/Free Full Text]
  29. Weidner N, Semple JP, Welch WR, Folkman J. Tumor angiogenesis and metastasis: correlation in invasive breast carcinoma. N Engl J Med 1991;324:1–8.[Abstract]
  30. Bartko J. On various intraclass correlation reliability coefficients. Psychol Bull 1976;83:762–765.[CrossRef]
  31. Shrout P, Fleiss J. Intraclass correlations: uses in assessing rater reliability. Psychol Bull 1979;86:420–428.[CrossRef]
  32. Fleiss J. The design and analysis of clinical experiments. New York, NY: Wiley, 1986.
  33. Libutti SK, Choyke P, Carrasquillo JA, Bacharach S, Neumann RD. Monitoring responses to antiangiogenic agents using noninvasive imaging tests. Cancer J Sci Am 1999;5:252–256.[Medline]
  34. Jain RK, Schlenger K, Hockel M, Yuan F. Quantitative angiogenesis assays: progress and problems. Nat Med 1997;3:1203–1208.[CrossRef][Medline]
  35. Graham CH, Rivers J, Kerbel RS, Stankiewicz KS, White WL. Extent of vascularization as a prognostic indicator in thin (< 0.76 mm) malignant melanomas. Am J Pathol 1994;145:510–514.[Abstract]
  36. Hollingsworth HC, Kohn EC, Steinberg SM, Rothenberg ML, Merino MJ. Tumor angiogenesis in advanced stage ovarian carcinoma. Am J Pathol 1995;147:33–41.[Abstract]
  37. Maeda K, Chung YS, Takatsuka S, et al. Tumour angiogenesis and tumour cell proliferation as prognostic indicators in gastric carcinoma. Br J Cancer 1995;72:319–323.[Medline]
  38. Takahashi Y, Kitadai Y, Bucana CD, Cleary KR, Ellis LM. Expression of vascular endothelial growth factor and its receptor KDR, correlates with vascularity, metastasis, and proliferation of human colon cancer. Cancer Res 1995;55:3964–3968.[Abstract/Free Full Text]
  39. Weidner N, Carroll PR, Flax J, Blumenfeld W, Folkman J. Tumor angiogenesis correlates with metastasis in invasive prostate carcinoma. Am J Pathol 1993;143:401–409.[Abstract]
  40. Iordanescu I, Becker C, Zetter B, Dunning P, Taylor GA. Tumor vascularity: evaluation in a murine model with contrast-enhanced color Doppler US effect of angiogenesis inhibitors. Radiology 2002;222:460–467.[Abstract/Free Full Text]
  41. Kim SW, Park SS, Ahn SJ, et al. Identification of angiogenesis in primary breast carcinoma according to the image analysis. Breast Cancer Res Treat 2002;74:121–129.[CrossRef][Medline]
  42. Huber S, Vesely M, Zuna I, Delorme S, Czembirek H. Fibroadenomas: computer-assisted quantitative evaluation of contrast-enhanced power Doppler features and correlation with histopathology. Ultrasound Med Biol 2001;27:3–11.[CrossRef][Medline]
  43. Strohmeyer D, Frauscher F, Klauser A, et al. Contrast-enhanced transrectal color doppler ultrasonography (TRCDUS) for assessment of angiogenesis in prostate cancer. Anticancer Res 2001;21:2907–2913.[Medline]
  44. Plunkett ML, Hailey JA. An in vivo quantitative angiogenesis model using tumor cells entrapped in alginate. Lab Invest 1990;62:510–517.[Medline]
  45. Eyding J, Wilkening W, Reckhardt M, et al. Contrast burst depletion imaging (CODIM): a new imaging procedure and analysis method for semiquantitative ultrasonic perfusion imaging. Stroke 2003;34:77–83.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 J Ultrasound MedHome page
D. J. Lee, A. Lyshchik, J. Huamani, D. E. Hallahan, and A. C. Fleischer
Relationship Between Retention of a Vascular Endothelial Growth Factor Receptor 2 (VEGFR2)-Targeted Ultrasonographic Contrast Agent and the Level of VEGFR2 Expression in an In Vivo Breast Cancer Model
J. Ultrasound Med., June 1, 2008; 27(6): 855 - 866.
[Abstract] [Full Text] [PDF]

Home page
 J Ultrasound MedHome page
A. Lyshchik, A. C. Fleischer, J. Huamani, D. E. Hallahan, M. Brissova, and J. C. Gore
Molecular Imaging of Vascular Endothelial Growth Factor Receptor 2 Expression Using Targeted Contrast-Enhanced High-Frequency Ultrasonography
J. Ultrasound Med., November 1, 2007; 26(11): 1575 - 1586.
[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 Lucidarme, O.
Right arrow Articles by Mattrey, R. F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lucidarme, O.
Right arrow Articles by Mattrey, R. F.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS