Noninvasive monitoring of antiangiogenic therapy was performed by serial power Doppler ultrasound imaging of murine tumors treated with recombinant interleukin 12, the results of which were correlated with assessments of tumor vascularity by microscopy. Growth of established K1735 tumors, but not of IFN-γ-unresponsive K1735.N23 variants, was suppressed by treatment. Serial Doppler imaging of K1735 tumor vascularity during treatment revealed a progressive change from a diffuse perfusion pattern to a more punctate distribution. Quantitative analysis of the images revealed that color-weighted fractional average, representing overall tumor perfusion, consistently decreased in these tumors, primarily because of a decrease in fractional tumor cross-sectional area carrying blood flow. In contrast, these parameters increased in nonresponsive tumors during treatment. Confocal microscopy of thick tumor sections revealed a reduction in the density and arborization of vessels labeled in vivo by fluorochrome-conjugated lectin with effective treatment. Immunohistological examination of thin tumor sections confirmed the preferential loss of small vessels with successful therapy. Similar changes in tumor vascular anatomy and perfusion were also observed during recombinant interleukin 12 treatment of two other responsive murine tumor types. These results indicate that power Doppler ultrasound is a sensitive, noninvasive method for reporting functional consequences of therapy-induced vascular anatomical changes that can be used to serially monitor tumor perfusion and efficacy of antivascular therapy in clinical trials.

Compelling evidence that tumor growth is angiogenesis dependent (reviewed in Ref. 1) has led to the testing of antiangiogenic or antivascular agents in cancer therapy. In mice, some of these agents have been shown to control tumor growth and produce long-lasting tumor regression (2, 3). Angiogenesis inhibitors do not act directly on tumor cells, but rather on the cells providing the tumor with its vasculature. This means that reduced tumor vascularity is the direct result of therapy, whereas control of tumor growth is a derivative therapeutic effect. As a consequence, clinical monitoring of tumor growth during antivascular therapy3 may not accurately or adequately reflect therapeutic efficacy. Ideally, an antivascular effect will result in tumor regression. However, the uncertain response of tumors to incomplete angiogenesis inhibition, together with the already highly unpredictable growth pattern of many human tumors, means that therapeutic success and failure will be difficult or impossible to distinguish on clinical grounds alone. Although theoretical models are being developed to predict tumor growth responses to antivascular therapy (4), knowing whether tumor vasculature is inhibited in these cases will be crucial for proper interpretation of results and helpful for design of therapeutic antivascular regimens. For these reasons, direct assessment of tumor vascularity and vascular function, as well as their change with therapy, may be important during evaluation of the many antivascular agents that promise to enter or have already entered clinical trials (5). In these trials, monitoring of tumor vascularity or parameters related to the vasculature, such as tumor oxygenation or perfusion, represents not only an intermediate end point of clinical efficacy but also the direct end point of therapeutic efficacy.

The method most often used to assess tumor vascular inhibition is immunohistological analysis of tumor MVD,4 which has been correlated with increased metastasis (6, 7, 8) and patient mortality (9, 10, 11) in a number of human tumor types. Treatment of animal tumors with angiogenesis inhibitors has been associated with a decrease in MVD (12, 13, 14). MVD measurements, however, may have limited utility in a clinical setting. The requirement for tissue to perform histological analysis presents a difficulty in performing analysis in patients, and because an MVD measurement reflects only a single point in time, multiple tissue specimens will probably be needed to establish a therapeutic effect. In addition, MVD provides an anatomical, but not a functional, measurement of tumor vascularity. Whereas many antiangiogenic therapies are likely to reduce vascularity by decreasing the number of vessels, they may also affect perfusion through vessels. This latter component may be particularly important for antivascular agents that have been shown to induce acute and massive tumor hemostasis (15, 16).

Other techniques for assessing changes in tumor vascularity will be needed. Ideally, such methods should be noninvasive to allow serial measurements of the same tumor mass in situ over time and should be sensitive to changes in functional tumor vascularity. One such method is Doppler ultrasound, which has been used extensively to assess vascularity in a range of animal and human tumors (reviewed in Refs. 17, 18, 19, 20, 21, 22). In particular, power (or amplitude) mode Doppler ultrasound, which color-encodes the integrated power of the Doppler signal rather than the conventional mean Doppler frequency shift, appears to be a more sensitive and specific method for measuring low-velocity tumor blood flow than traditional color Doppler ultrasound (23, 24, 25). In this study, we examined power Doppler ultrasound as a potential method for noninvasive monitoring of tumor vascular inhibition. In addition, we explored changes in tumor vasculature during therapy and correlated them with the changes detected by ultrasound.

Mice and Cell Lines.

C57BL/6 and BALB/c mice were purchased from The Jackson Laboratory (Bar Harbor, ME); C3H/HeN mice were purchased from Harlan Sprague Dawley (Indianapolis, IN). All mice were 6–8-week-old females maintained in microisolator cages under sterile conditions. The K1735 (26) and B16F10 (27) murine melanoma cell lines (syngeneic with C3H/HeN and C57BL/6 strains, respectively), along with the RENCA (28) renal cell adenocarcinoma cell line (syngeneic with BALB/c), were maintained in DMEM supplemented with 10% FCS and penicillin/streptomycin. Generation of stably transfected N23 K1735 cells containing a dominant-negative IFNγR1, which renders them unresponsive to the antitumor effects of rmIL-12, was described previously (29).

In Vivo Studies.

For tumor growth studies, 106 tumor cells were injected s.c. into the lower left flank of syngeneic mice. Injected cells were derived from low-passage frozen stocks that had been established in culture less than 1 week prior to injection. When established tumors reached a diameter of 3–4 mm, rmIL-12 or PBS vehicle was administered i.p. at the maximum tolerated dose for each strain on a five dose per week schedule (five daily injections followed by 2 days of rest) for up to 3 weeks. C3H/HeN mice received 125 ng per injection, whereas C57BL/6 and BALB/c mice received 500 ng per injection. Tumors were measured bidirectionally by calipers at regular intervals, and tumor volume was calculated using the formula for approximating the volume of a spheroid: 0.52 × (width)2 × (length). Mice were euthanized according to guidelines established by the Institutional Animal Care and Use Committee.

Doppler Ultrasound Imaging of Tumors.

Power Doppler ultrasound imaging of tumors was performed essentially according to a previously published method (18). Tumor-bearing mice were anesthetized (140 mg/kg ketamine, 1.3 mg/kg xylazine) to minimize echogenicity attributable to tissue motion. After haircoat clipping in the area overlying the tumor, the mice were placed in sternal recumbency to facilitate tumor alignment with the ultrasound transducer. Imaging was performed using an Ultramark 9 HDI ultrasound machine (Advanced Technology Laboratories, Inc., Bothell, WA) with a L10-5 MHz transducer. A 5-mm acoustic standoff between transducer face and tumor was achieved by generous application of acoustic gel. Power Doppler measurement settings were held constant for all tumors (80% color gain; 50-Hz wall filter; 600-Hz pulse repetition frequency). Imaging lasted ∼5–10 min per mouse, with each mouse being anesthetized immediately before imaging and placed on a warm pad to minimize changes in body temperature. Each tumor was imaged in its entirety at serial 1-mm intervals through the longest axis. Initial scanning of each tumor was performed in B-mode (grayscale ultrasound) to define the boundary of the tumor mass based on echogenicity parameters. A rectangular area was then placed around the tumor and surrounding tissue, denoting the region in which power Doppler data would be acquired. The integrated power values of the Doppler spectrum from corpuscular flow were displayed visually on a scale that converted power values to color hue and saturation levels (22). Images were recorded on videotape (S-VHS format) and digitized frame by frame at 24-bit resolution using a Macintosh AV-7600 frame grabber.

For analysis, a ROI was drawn around the tumor boundary based on the initial B-mode scan. Three measurements were obtained for each ROI: MCL, FA, and CWFA. The MCL was obtained by dividing the sum of the integrated power values by the number of colored pixels. The FA was the ratio of colored pixels to the total number of pixels in the ROI. The product of the first two parameters determined CWFA. Blood flow parameters (22) were determined for each tumor by individual quantitation of parameters for three or more images within each series followed by averaging of individual values.

Confocal Microscopy of Tumor Vasculature.

Tumor-bearing mice received i.v. injections of 150 μl of 1 mg/ml FITC-conjugated tomato (Lycopersicon esculatum) lectin (Vector Labs, Burlingame, CA) in PBS into the tail vein 15 min prior to tumor excision. Following excision, the tumors were sectioned manually into thick (0.5–1.0 mm) slices that were mounted onto microscope slides with 50% glycerol in PBS and covered with a coverslip. Slides were examined using an upright Nikon (Augusta, GA) E-600 Eclipse microscope equipped with a Bio-Rad (Hercules, CA) 1024-ES confocal system. FITC fluorescence was detected by a three-line, 15-mW Argon-Krypton laser system (American Laser, Fraser, MI). Images were viewed by ×10 objective lens with field dimensions of 1004.5 × 1004.5 μm. For each slide, serial images were acquired at 2.5-μm intervals over a standard 100-μm depth, using Bio-Rad Lasersharp Acquisition software, and integrated to create a composite maximum intensity projection of tumor vasculature imaged in three dimensions. Projection images were analyzed using ImageTool software (University of Texas, San Antonio, TX) for vessel density, luminal diameter, and arborization. The vessel density of each image was defined as the number of vessel intersections with a four-axis grid (vertical, horizontal, and two diagonal axes through center of the image) superimposed over the image. Lumen cross-sectional diameters were determined for all intersecting vessels. A vessel arborization index was then determined for each image by dividing the total number of vessel branchpoint nodes within the image volume by the number of vessels.

Immunohistochemistry.

Thin sections (4 μm) from formalin-fixed, paraffin-embedded tumors were stained for the endothelial cell marker vWF. Tissue slides were deparaffinized and incubated in 0.3% hydrogen peroxide for 10 min at 4°C. Antigen retrieval was performed by incubation in 0.12% Pronase (Boehringer Mannheim; Indianapolis, IN) for 15 min at 37°C followed by blocking with PBS containing 0.1% BSA and 5% goat serum for 20 min at 37°C. The tissue was then stained with a polyclonal rabbit anti-vWF antibody (Dako, Carpinteria, CA) diluted 1:1500 in blocking solution for 2 h at room temperature. Slides were then incubated with biotinylated goat antirabbit immunoglobulin antibody (Vector Labs) diluted 1:200 in blocking solution for 1 h at room temperature. Slides were incubated in streptavidin-horseradish peroxidase (Research Genetics, Huntsville, AL) for 1 h at room temperature and subsequently developed using amino ethyl carbazole substrate (Vector Labs). Slides were then counterstained with hematoxylin.

Image Acquisition and Analysis.

All histological specimens were viewed under a Nikon light microscope equipped with a Hawamatsu digital camera and Nikon ImagePro acquisition software. Images were analyzed using ImageTool software. For MVD measurements, slides were scanned at low power (×40 magnification) to identify areas of highest vascularity. Twenty high-power (×20x) fields were then selected randomly within these areas, and MVDs were calculated based on the number of vWF-positive structures. In addition, vessel lumen cross-sectional areas were determined for all counted vessels automatically based on spatial calibration parameters established with a slide micrometer. Microvessel were counted by multiple blinded observers in conjunction with a pathologist. Three sections per tumor were analyzed from five tumors within each group.

Statistical Analysis.

Assessment of statistical significance was performed either by Student’s t test (for normally distributed data sets) or the Mann-Whitney U test (for nonnormally distributed data sets). Correlation coefficients were derived from Pearson’s correlation method. All statistical analysis was performed using Instat software for the Macintosh, version 2.0 (Graphpad Software; Philadelphia, PA).

Growth of K1735 but not K1735.N23 Tumors Is Suppressed by rmIL-12 Therapy.

Angiogenesis and growth of K1735 tumors both have been shown previously to be suppressed by rmIL-12 therapy (30). When these tumors are engineered to overexpress a dominant-negative IFN-γ receptor R1 [K1735.N23 tumors from Coughlin et al.(29)], they become unresponsive to both the antiangiogenic and antitumor effects of rmIL-12 (30). In this study, we used K1735 and K1735.N23 tumors treated with rmIL-12 as examples of effective and ineffective antiangiogenesis therapy, respectively. Representative of the results of numerous such studies, Fig. 1,A shows the growth of s.c. K1735 tumors treated with rmIL-12 after they reached 3–4 mm in diameter. Growth of four of these (Fig. 1A, K1–K4) was markedly suppressed compared with growth of the untreated, concurrent controls (Fig. 1,A, CTRL, which is the average of 10 untreated tumors). One tumor (Fig. 1,A, K5) was poorly controlled by treatment. This was highly unusual for K1735 tumors treated with rmIL-12 and warranted further investigation. In contrast, growth of five K1735.N23 tumors treated with rmIL-12 was indistinguishable from the control tumors (Fig. 1 B).

rmIL-12 Suppression of Tumor Growth Is Associated with Progressive Reduction in Tumor Blood Flow Measured by Power Doppler Ultrasound.

We assessed the vascularity of tumors noninvasively by power Doppler ultrasound to correlate changes in blood flow with alterations in growth during therapy. For each tumor imaged, three blood flow parameters were measured: MCL, which represents the number of RBCs causing the Doppler shift per unit time (RBC flux); FA, the proportion of the cross-sectional tumor area carrying flow; and CWFA (MCL × FA), which offers an overall measurement of tumor perfusion. A progressive decrease in tumor perfusion was observed in K1735 tumors that responded to rmIL-12 treatment (Fig. 2). Although a decrease was seen in one tumor after only 1 week of treatment, all responding tumors exhibited a detectable decrease after 3 weeks of treatment. Visually, the pattern of blood flow in these tumors changed from a diffuse to a patchy distribution.

Quantitative analysis of the Doppler images revealed that rmIL-12 treatment led to a dramatic decrease in CWFA in K1735 tumors (Fig. 3,A). The decrease in CWFA was seen despite a gradual increase in size of the K1735 tumors during treatment (Fig. 1,A) and primarily reflected a decrease in FA (Fig. 3,B; both parameters were on average 35–40% below pretreatment values at week 3 of therapy), whereas MCL levels changed little. Therapeutically unresponsive K1735.N23 tumors, in contrast, exhibited consistent increases in CWFA and FA during the course of rmIL-12 treatment (average increases of 23 and 19%, respectively). These increases were consistent with levels seen in size-matched, untreated K1735 tumors (Fig. 3,A). Thus, power Doppler ultrasound is a sensitive method for detecting changes in tumor blood flow within tumors undergoing a therapeutic response. Significantly, no decrease in either CWFA or FA was observed in K1735 tumor K5 (Fig. 3), which continued to grow at a near normal rate despite rmIL-12 treatment. This observation may explain why this tumor was unresponsive to therapy. Solely on the basis of tumor growth, one could not say whether this tumor failed to respond because of a failure to reduce tumor perfusion or because the tumor was relatively resistant to the effects of vascular inhibition: power Doppler ultrasound revealed that the former is the likely explanation.

Decreased Blood Flow Is Attributable to Decreases in Density and Arborization of Perfused Vessels.

rmIL-12 is known to inhibit tumor cell-mediated angiogenesis (30, 31), so we characterized the antivascular effect of rmIL-12 within tumors to provide a biological correlate for the decrease in blood flow observed by power Doppler ultrasound. We injected a fluorescently conjugated tomato lectin (32), which binds to carbohydrate groups on the surface of endothelial cells, i.v. into mice bearing size-matched untreated and rmIL-12-treated K1735 tumors and examined functional tumor vessels (i.e., those with access to lectin) in three dimensions by serial confocal scanning microscopy. This method revealed that rmIL-12 markedly reduced vessel density and branching in K1735 tumors (Fig. 4,A) and provided direct visual evidence of its antivascular effect in treated tumors. There was an overall decrease (Fig. 4,B) in the density of perfused vessels per ×10 objective tumor volume (0.1 mm3) with treatment (37.0 ± 9.9 untreated versus 22.3 ± 4.0 treated) that was associated with a reduction in vessel arborization (1.94 ± 0.23 nodes per vessel in untreated tumors versus 1.01 ± 0.23 nodes per vessel in treated tumors) and an increase in average vessel diameter (25.6 ± 5.1 μm in treated tumors compared with 14.8 ± 2.32 μm in untreated tumors). All three of these changes in tumor vascularity were found to be statistically significant (P < 0.05, Mann-Whitney U test). This increase in vessel diameter with treatment may partly explain changes in the overall vascular pattern observed by power Doppler ultrasound, where an evenly distributed perfusion pattern pretherapy is replaced by a more patchy and focal pattern posttherapy (Fig. 2). Examination of confocal images taken from serial thick tumor sections demonstrated that these effects of rmIL-12 were not confined to a particular region of the tumor but appeared throughout. K1735.N23 tumors did not undergo any significant alteration in tumor vessel density, size, or branching with treatment (data not shown).

rmIL-12 Treatment Reduces MVD through Selective Ablation of Small-Caliber Vessels.

Tumor vascularity was then assessed by immunohistological staining of thin (4 μm) sections for the endothelial marker vWF followed by MVD measurement, a standard method for assessing tumor vascularity. We compared sections taken from size-matched K1735 and K1735.N23 tumors that were either untreated or treated with rmIL-12 (Fig. 5,A). The treated tumors examined were the same tumors monitored by power Doppler ultrasound. When these stained sections were analyzed for their density of vWF-positive structures, rmIL-12 treatment was found to decrease the density of vessels in K1735 tumors from 11.6 ± 2.3 vessels/high-power field to 6.6 ± 0.56 (Fig. 5,B; P < 0.01). Only results from tumors K1–K4, which responded to rmIL-12 treatment, were included in this analysis of the treated group. MVD of K1735.N23 tumors (10.1 ± 1.9 untreated versus 8.4 ± 0.32 treated) was not significantly altered with therapy (P > 0.05). We also examined MVD in tumor K5 separately and found that its MVD (10.9 ± 2.2) was not significantly different from that of the untreated group. This was consistent with the lack of reduction in tumor blood flow seen in this tumor by power Doppler ultrasound (Figs. 2,B and FIG 3 A). Thus, the two methods for assessing tumor vascularity, confocal imaging and MVD determination, each indicated a level of vascular decrease (40 and 46%, respectively) with rmIL-12 therapy that was consonant with the decrease in tumor perfusion (37%) indicated by power Doppler ultrasonography.

To further correlate the histological pattern of K1735 vessel regression with changes seen by power Doppler ultrasound and confocal microscopy, we obtained computer-generated measurements of vessel lumen cross-sectional areas on the vWF-stained histological sections of size-matched rmIL-12-treated and untreated K1735 tumors (Fig. 6). A comparison of vessel size distribution by this method revealed a dramatic reduction (63%) in vessels with cross-sectional areas <200 μm2, whereas all other vessels remained essentially unchanged (3% increase overall with treatment). This decrease in the number of small vessels likely represents the loss of small-caliber vessel branches seen by confocal microscopy and is responsible for the decreases in CWFA and FA seen by power Doppler ultrasound.

Power Doppler Ultrasound Can Detect rmIL-12-induced Blood Flow Reduction in Other Mouse Tumors.

Having established the utility of power Doppler ultrasound for noninvasive monitoring of angiogenesis inhibition in K1735, we tested its utility in other tumors. Treatment of both B16F10 melanomas and RENCA renal cell carcinomas with rmIL-12 led to marked suppression of tumor growth (Fig. 7,A). Serial power Doppler ultrasound imaging of these tumors during therapy revealed a reduction in tumor blood flow similar to that seen in K1735 tumors (Fig. 7,B). B16F10 and RENCA tumors, in fact, exhibited more dramatic decreases in blood flow with treatment (Fig. 7,C). Average CWFA decreased 57 and 79%, respectively, at week 3 of treatment versus pretreatment time points. Like K1735 tumors, the drop in CWFA observed in these two other tumors was primarily attributable to a decrease in FA, with MCL changing minimally (Fig. 7 C). To correlate these observations with changes in tumor vascularity, MVD was assessed in these other two tumor types in size-matched rmIL-12-treated or untreated tumors. This revealed that both B16F10 and RENCA tumors exhibited statistically significant decreases in MVD after 3 weeks of treatment (from 9.87 ± 1.73 to 3.99 ± 2.03 and from 9.45 ± 0.87 to 6.79 ± 0.68 vessels/high-power field, respectively; P < 0.005, Student’s t test), providing an anatomical correlate for the power Doppler ultrasound measurements.

This study evaluated Doppler ultrasound as a method for assessing tumor vascular inhibition. If Doppler ultrasound can detect therapeutic alterations in tumor vascularity, its widespread availability and use in clinical radiology, together with its noninvasiveness, make this method a convenient and immediately applicable way to determine the efficacy of antivascular agents in clinical trials. Doppler ultrasound has been studied as a monitoring tool for antivascular therapy (19, 21), but the changes in perfusion detected by ultrasound have never been linked to a change in tumor vascularity. This study used power (amplitude) mode Doppler ultrasound, which has been shown to be a more sensitive and specific technique for measuring low-velocity blood flow than traditional color Doppler ultrasound because of the low noise variance and increased dynamic range of power Doppler (23). The tumors monitored varied in size from ∼3 to 12 mm in diameter, which indicates the applicability of this method for even small tumors.

We serially monitored s.c. K1735 and K1735.N23 tumors during rmIL-12 treatment by power Doppler ultrasound as examples of tumors that are susceptible and resistant to antivascular therapy, respectively, and found a strong correlation between therapeutic reduction of tumor growth and reduction in tumor blood flow. Overall blood flow reflected in the CWFA decreased by ∼46% following rmIL-12 therapy compared with pretreatment levels in K1735 tumors, which exhibited growth retardation. In contrast, CWFA levels increased by 23% in therapeutically nonresponsive K1735.N23 tumors, which continued to grow normally during rmIL-12 treatment. A similar increase in blood flow with size was observed in untreated K1735 tumors, and although we do not have a clear explanation for this, it is clear that the tumor vasculature is able to keep pace with tumor growth. Consistent with this, K1735 tumors do not exhibit severe tumor cell hypoxia or spontaneous necrosis, even at large sizes (Ref. 30 and data not shown). Quantitative analysis revealed that the decrease in CWFA in the responding tumors was almost exclusively attributable to a decrease in the average tumor cross-sectional area bearing flow (FA), whereas mean red cell flux (MCL) remained unchanged. In the unusual K1735 tumor that grew normally despite therapy, power Doppler ultrasound yielded blood flow values comparable to those of size-matched, untreated tumors. This lends weight to both the sensitivity and specificity of power Doppler as a method for assessing therapeutic vascular inhibition. The results in K1735 tumors were confirmed in two other mouse tumor types, B16F10 and RENCA, which differ with respect to mouse genetic background and baseline vascularity. Thus power Doppler ultrasound is a generally effective tool for monitoring tumor perfusion.

Theoretically, Doppler ultrasound should also be sensitive to tumor vascular changes arising from therapeutic alterations in MCL rather than FA. This might be observed in tumors undergoing antivascular therapy with an agent such as combretastatin, which has been shown to reduce tumor blood flow within hours of therapy, well before vessels would have regressed (15, 16, 33). Although such studies have not yet been performed, the potential for Doppler ultrasound to yield insight into different mechanisms of vascular regression should be further explored.

An important goal of this study was to characterize the vascular changes that give rise to the changes in blood flow detect by power Doppler ultrasound. Given that the decrease in CWFA was primarily attributable to a drop in FA, we wanted to determine whether this was the result of a reduction in the size or in the number of functional blood vessels. Confocal microscopic analysis revealed that there was a reduction in the density of perfused blood vessels with treatment accompanied by a significant decrease in the degree of vessel arborization and an increase in average vessel size. This pattern of vascular inhibition was also confirmed by histological staining of thin sections for endothelial cells and microvessel analysis. It is interesting to note that rmIL-12 treatment of K1735 tumors, in addition to causing regression of small vessels, also was associated with an increase in caliber of the remaining vessels (see Fig. 4,A and vessel categories above 500 μm2 in Fig. 6). Although we do not have a good explanation for this observation, the dilated appearance of tumor vessels in the treated tumors may be related to a compensatory increase in flow through remaining vessels following therapeutic vessel regression. The reduction in vessel density is attributable, overall, to a reduction in vessel branching, which is consistent with inhibition of sprouting angiogenesis. We found a statistically significant (P < 0.003, t test) positive correlation between Doppler blood flow and MVD measurements (r = 0.80 for MVD versus FA; r = 0.83 for MVD versus CWFA), which provides a link between blood vessel anatomy and functionality. The ability of clinical frequency Doppler ultrasound to detect flow through tumor microvessels is the subject of active investigation. Although the sensitivity of the technique is certainly dependent on both vessel caliber as well as flow rate, previous studies have demonstrated Doppler detection of flow through tumor vessels as small as 10–20 μm in diameter (34, 35). We believe that loss of small vessels (<300 μm2) is largely responsible for the reduction in FA that underlies the drop in CWFA. This link between vessel anatomy and functionality provides justification for using methods, such as MVD, that quantitate anatomical vessels to assess tumor vascularity and for using reduction in vessel density as an indicator of effective antivascular therapy.

The noninvasive nature of Doppler ultrasound imaging is a major advantage for its use as a clinical monitoring tool. It measures vascularity of tumors in situ without perturbing the tumor being studied, which allows serial assessment of the same tumor mass. This is particularly advantageous given the variability in perfusion among tumors. We have observed significant variability in CWFA measurements between different tumor types by power Doppler ultrasound (e.g., perfusion parameters in untreated K1735 and RENCA tumors can be equivalent to those of a B16F10 tumor in the third week of rmIL-12 treatment). In addition, tumors of the same type in different mice may have different patterns and values of perfusion at baseline (note the initial pretreatment CWFA values in Fig. 3 A). This indicates that serial, rather than single, measurements of tumor perfusion obtained from the same tumor likely will be needed to determine effectiveness of antivascular therapy. Of course, the reproducibility of any noninvasive technique will be critical to validate comparison of serial measurements from the same tumor. To this end, we performed power Doppler imaging on a number of K1735 tumors at multiple time points spaced several minutes apart to assess the reproducibility of the technique and found that average tumor CWFA measurements for each tumor varied by ≤10% over the range of points (data not shown). Another important consideration for these measurements is that they encompass the entire tumor to avoid regional variations in flow within each tumor. This would allow fair comparison of serial flow assessments. In this regard, we assessed global tumor vascularity by performing a sweep of the entire tumor with the ultrasound transducer and capturing images sequentially through the long axis of the tumor. Overall measurements were then derived from the sum total of these captured images.

The development of clinical methods for assessing tumor vascular inhibition will become increasingly important as antivascular agents enter human trials. In this and most other mouse studies, tumor response to therapy was easy to discern because of the predictable growth of the tumors under study and the availability of controls. In contrast, the highly variable (often fitful) growth of many human tumors and the absence of control tumors in cancer patients for comparison make determination of therapeutic response in human clinical trials problematic. This points to the importance of assessing tumor vascular response to therapy for proper evaluation of therapeutic antivascular regimens. Our results illustrate this point. rmIL-12-treated K1735 tumors generally do not regress during therapy, instead exhibiting slow increases in size. The progressive decreases in power Doppler CWFA and FA levels with treatment provide unequivocal evidence of therapeutic response, obviating the need to rely on tumor growth kinetics.

In conclusion, our studies show that power Doppler ultrasound is a sensitive and specific noninvasive method for monitoring of tumor perfusion and vascularity. There may be factors we have not explored that potentially limit the ability of this technology to measure tumor perfusion (e.g., location and echogenicity of the tumor and its surrounding tissues), but alternative imaging modalities (e.g., magnetic resonance imaging) may be useful in these cases (36, 37). In addition, although ultrasonography does not directly yield mechanistic information about therapeutic effects on blood vessel anatomy, our accompanying histological studies indicate functional and anatomical vascular correlates that shed mechanistic light. With little question, the ability of power Doppler ultrasound to monitor tumor perfusion conveniently, safely, and repeatedly recommends its incorporation into clinical trials of antivascular therapies. With the information gained about effects of agents on this important intermediate therapeutic end point, development of these therapies for cancer may proceed along more rational lines, and their full potential may be more quickly realized.

Fig. 1.

rmIL-12 effects on K1735 (A) and K1735.N23 (B) tumors. Syngeneic C3H/HeN mice were inoculated with 106 K1735 or K1735.N23 (an IFN-γ-unresponsive K1735 derivative that is resistant to the antitumor effects of IL-12) cells s.c. When tumors reached 3–4 mm in diameter, rmIL-12 or PBS injections were initiated on a five injection/week schedule for 3 weeks. Each panel depicts tumor volumes of vehicle-treated tumors (▪; average of 10 tumors; bars, SD) and individual rmIL-12-treated (K1–K5 in A and N1–N5 in B) tumors at 2-day intervals. The arrow indicates the day on which rmIL-12 or PBS administration was initiated.

Fig. 1.

rmIL-12 effects on K1735 (A) and K1735.N23 (B) tumors. Syngeneic C3H/HeN mice were inoculated with 106 K1735 or K1735.N23 (an IFN-γ-unresponsive K1735 derivative that is resistant to the antitumor effects of IL-12) cells s.c. When tumors reached 3–4 mm in diameter, rmIL-12 or PBS injections were initiated on a five injection/week schedule for 3 weeks. Each panel depicts tumor volumes of vehicle-treated tumors (▪; average of 10 tumors; bars, SD) and individual rmIL-12-treated (K1–K5 in A and N1–N5 in B) tumors at 2-day intervals. The arrow indicates the day on which rmIL-12 or PBS administration was initiated.

Close modal
Fig. 2.

Serial power Doppler ultrasound imaging of tumors during rmIL-12 therapy. K1735 and K1735.N23 tumors were imaged at weekly intervals during therapy to determine tumor blood flow. A, serial images from two K1735 tumors at time points just before therapy initiation (week 0), after 1 week (week 1) and after 3 weeks (week 3) of rmIL-12 therapy. Each set of images is from the same tumor. B, comparison of Doppler ultrasound images taken from individual tumors (panels labeled in upper right-hand corner; K, K1735; N, N23) that were either responsive or unresponsive to rmIL-12 antitumor activity (see Fig. 1 for tumor growth kinetics). The images were taken at the therapeutic time points indicated. Each pair of images is taken from the same tumor. For each image, a dotted line demarcates the tumor ROI, and regions of blood flow are depicted in color. The boxed area indicates the total field in which ultrasound data were acquired.

Fig. 2.

Serial power Doppler ultrasound imaging of tumors during rmIL-12 therapy. K1735 and K1735.N23 tumors were imaged at weekly intervals during therapy to determine tumor blood flow. A, serial images from two K1735 tumors at time points just before therapy initiation (week 0), after 1 week (week 1) and after 3 weeks (week 3) of rmIL-12 therapy. Each set of images is from the same tumor. B, comparison of Doppler ultrasound images taken from individual tumors (panels labeled in upper right-hand corner; K, K1735; N, N23) that were either responsive or unresponsive to rmIL-12 antitumor activity (see Fig. 1 for tumor growth kinetics). The images were taken at the therapeutic time points indicated. Each pair of images is taken from the same tumor. For each image, a dotted line demarcates the tumor ROI, and regions of blood flow are depicted in color. The boxed area indicates the total field in which ultrasound data were acquired.

Close modal
Fig. 3.

Quantitative analysis of tumor blood flow in rmIL-12-treated K1735 and K1735.N23 tumors. A set of power Doppler images was taken from each tumor at weekly intervals during rmIL-12 therapy and analyzed for three parameters: overall tumor perfusion (CWFA), cross-sectional tumor area bearing flow (FA), and average red cell flux (MCL). Tumor blood flow parameters are plotted against tumor volume at the time of imaging. A, CWFA measurements for treated K1735 and K1735.N23 tumors. B, FA and MCL data from K1735 tumors. ▪ depict data for untreated tumors as a comparison; other symbols represent serial measurements of individual treated tumors (dashed lines connect serial measurements of the same tumor). For K1735 tumors (K1–K5), dashed lines connect measurements obtained at weeks 0, 1, 2, and 3 of rmIL-12 treatment, or weeks 0 and 3 (for tumors 3 and 4). For K1735.N23 tumors (N1–N5), dashed lines connect measurements obtained at weeks 0 and 2 of rmIL-12 therapy.

Fig. 3.

Quantitative analysis of tumor blood flow in rmIL-12-treated K1735 and K1735.N23 tumors. A set of power Doppler images was taken from each tumor at weekly intervals during rmIL-12 therapy and analyzed for three parameters: overall tumor perfusion (CWFA), cross-sectional tumor area bearing flow (FA), and average red cell flux (MCL). Tumor blood flow parameters are plotted against tumor volume at the time of imaging. A, CWFA measurements for treated K1735 and K1735.N23 tumors. B, FA and MCL data from K1735 tumors. ▪ depict data for untreated tumors as a comparison; other symbols represent serial measurements of individual treated tumors (dashed lines connect serial measurements of the same tumor). For K1735 tumors (K1–K5), dashed lines connect measurements obtained at weeks 0, 1, 2, and 3 of rmIL-12 treatment, or weeks 0 and 3 (for tumors 3 and 4). For K1735.N23 tumors (N1–N5), dashed lines connect measurements obtained at weeks 0 and 2 of rmIL-12 therapy.

Close modal
Fig. 4.

Confocal microscopy of tumor vasculature. Mice bearing either untreated or rmIL-12 treated K1735 tumors received injections of a FITC-conjugated lectin prior to tumor excision. Thick tumor sections were then cut and imaged by confocal microscopy. A, representative composite projections of serial confocal image sets taken over a 100-μm range at ×10 magnification for size-matched treated and untreated tumors. Scale bar, 100 μm. B, quantitative analysis of composite images from untreated (□) and treated () tumors for density of perfused vessels (left), average vessel diameter (middle), and degree of vessel branching (right). Bars, SD; *, statistically significant difference (P < 0.05, Mann-Whitney U test).

Fig. 4.

Confocal microscopy of tumor vasculature. Mice bearing either untreated or rmIL-12 treated K1735 tumors received injections of a FITC-conjugated lectin prior to tumor excision. Thick tumor sections were then cut and imaged by confocal microscopy. A, representative composite projections of serial confocal image sets taken over a 100-μm range at ×10 magnification for size-matched treated and untreated tumors. Scale bar, 100 μm. B, quantitative analysis of composite images from untreated (□) and treated () tumors for density of perfused vessels (left), average vessel diameter (middle), and degree of vessel branching (right). Bars, SD; *, statistically significant difference (P < 0.05, Mann-Whitney U test).

Close modal
Fig. 5.

Tumor MVD measurements. Histological sections from untreated and rmIL-12-treated (size-matched) K1735 and K1735.N23 tumors were stained for vWF, and the number of vWF-positive vessels (indicated by arrowheads) per high-power field was determined. A, representative high-power fields from untreated (left) and rmIL-12-treated (right) K1735 tumors. Scale bar, 50 μm. Note the reduced number and increased caliber of blood vessels (indicated by arrowheads) in the treated tumor relative to the untreated one. B, summary of average MVD measurements from untreated ▪) and rmIL-12-treated () tumors. Three sections were analyzed per tumor, with 15 high-power fields acquired within each section. Bars, SE; *, indicate statistical significance by Student’s t test (P < 0.01).

Fig. 5.

Tumor MVD measurements. Histological sections from untreated and rmIL-12-treated (size-matched) K1735 and K1735.N23 tumors were stained for vWF, and the number of vWF-positive vessels (indicated by arrowheads) per high-power field was determined. A, representative high-power fields from untreated (left) and rmIL-12-treated (right) K1735 tumors. Scale bar, 50 μm. Note the reduced number and increased caliber of blood vessels (indicated by arrowheads) in the treated tumor relative to the untreated one. B, summary of average MVD measurements from untreated ▪) and rmIL-12-treated () tumors. Three sections were analyzed per tumor, with 15 high-power fields acquired within each section. Bars, SE; *, indicate statistical significance by Student’s t test (P < 0.01).

Close modal
Fig. 6.

Histological determination of tumor vessel size. Size-matched untreated and rmIL-12-treated K1735 tumors were sectioned and stained for vWF to visualize blood vessels. High-power fields (hpf) were then imaged digitally and analyzed for vessel lumen cross-sectional area using automated software. Fig. 6 is a histogram showing the frequency of vessels per high-power field as a function of vessel size (in 100-μm2 intervals). The results represent the sum total of at least five tumors per group. Images were acquired in the same manner used for MVD analysis.

Fig. 6.

Histological determination of tumor vessel size. Size-matched untreated and rmIL-12-treated K1735 tumors were sectioned and stained for vWF to visualize blood vessels. High-power fields (hpf) were then imaged digitally and analyzed for vessel lumen cross-sectional area using automated software. Fig. 6 is a histogram showing the frequency of vessels per high-power field as a function of vessel size (in 100-μm2 intervals). The results represent the sum total of at least five tumors per group. Images were acquired in the same manner used for MVD analysis.

Close modal
Fig. 7.

Comparison of power Doppler ultrasound imaging of multiple tumor types treated with rmIL-12. A, effects of rmIL-12 treatment on the growth of established B16F10 melanomas (left) and RENCA renal cell adenocarcinomas (right). Each line depicts average tumor volumes as measured at 2-day intervals; bars, SD. B, representative serial Doppler ultrasound images taken from a B16F10 (left) and a RENCA (right) tumor at week 0 (top) and week 3 (bottom) of rmIL-12 therapy. The color intensity scale has been rendered in grayscale for display purposes. C, summary of the change in power Doppler ultrasound measurements of CWFA, FA, and MCL with rmIL-12 treatment for all of the tumors analyzed. Each column represents the average change in each parameter from week 0 to week 3 of therapy as determined by serial imaging. Bars, SD.

Fig. 7.

Comparison of power Doppler ultrasound imaging of multiple tumor types treated with rmIL-12. A, effects of rmIL-12 treatment on the growth of established B16F10 melanomas (left) and RENCA renal cell adenocarcinomas (right). Each line depicts average tumor volumes as measured at 2-day intervals; bars, SD. B, representative serial Doppler ultrasound images taken from a B16F10 (left) and a RENCA (right) tumor at week 0 (top) and week 3 (bottom) of rmIL-12 therapy. The color intensity scale has been rendered in grayscale for display purposes. C, summary of the change in power Doppler ultrasound measurements of CWFA, FA, and MCL with rmIL-12 treatment for all of the tumors analyzed. Each column represents the average change in each parameter from week 0 to week 3 of therapy as determined by serial imaging. Bars, SD.

Close modal

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work is supported by an NIH Medical Scientist Training Program grant (to M. S. G.) and NIH Grants RO1 CA74071 (to W. T. J., H. M. S., and S. M. E.), CA77851, and CA83042 (to W. M. F. L.).

3

We use the term “antivascular therapy” rather than “antiangiogenesis therapy” to encompass therapies that may have effects on preformed tumor vessels as well as newly forming vessels.

4

The abbreviations used are: MVD, microvessel density; rmIL-12, recombinant murine interleukin 12; ROI, region of interest; MCL, mean color level; FA, fractional area; CWFA, color-weighted fractional average; vWF, von Willebrand factor.

We would like to thank Genetics Institute (Andover, MA) for providing rmIL-12, the immunohistochemistry lab in the Pathology Department at the Hospital of the University of Pennsylvania for advice and reagents, and Dr. Newman Yeilding for helpful comments and critical reading of the manuscript. We also gratefully acknowledge the Hospital of the University of Pennsylvania GI Morphology Core Laboratory (NIH P30 DK 50306), where all histological images were acquired.

1
lkman J. What is the evidence that tumors are angiogenesis dependent?.
J. Natl. Cancer Inst. (Bethesda)
,
82
:
4
-6,  
1990
.
2
Boehm T., Folkman J., Browder T., O’Reilly M. S. Antiangiogenic therapy of experimental cancer does not induce acquired drug resistance.
Nature (Lond.)
,
390
:
404
-407,  
1997
.
3
O’Reilly M. S., Holmgran L., Chen C., Folkman J. Angiostatin induces and sustains dormancy of human primary tumors in mice.
Nat. Med.
,
2
:
689
-692,  
1996
.
4
Hahnfeldt P., Panigrahy D., Folkman J., Hlatky L. Tumor development under angiogenic signaling: a dynamic theory of tumor growth, treatment response, and postvascular dormancy.
Cancer Res.
,
59
:
4770
-4775,  
1999
.
5
Nelson N. J. Inhibitors of angiogenesis enter phase III testing.
J. Natl. Cancer Inst. (Bethesda)
,
90
:
960
-963,  
1998
.
6
Weidner N., Semple J. P., Welch W. R., Folkman J. Tumor angiogenesis and metastasis-correlation in invasive breast carcinoma.
N. Engl. J. Med.
,
324
:
1
-8,  
1991
.
7
Weidner N., Carroll P. R., Flax J., Blumenfeld W., Folkman J. Tumor angiogenesis correlates with metastasis in invasive prostate carcinoma.
Am. J. Pathol.
,
143
:
401
-409,  
1993
.
8
Fox S. B., Leek R. D., Bliss J., Mansi J. L., Gusterson B., Gatter K. C., Harris A. L. Association of tumor angiogenesis with bone marrow micrometastases in breast cancer patients.
J. Natl. Cancer Inst. (Bethesda)
,
89
:
1044
-1049,  
1997
.
9
Li V. W., Folkerth R. D., Watanabe H., Yu C., Rupnick M., Barnes P., Scott R. M., Black P. M., Sallan S. E., Folkman J. Microvessel count and cerebrospinal fluid basic fibroblast growth factor in children with brain tumors.
Lancet
,
344
:
82
-86,  
1994
.
10
Weidner N., Folkman J., Pozza F., Bevilacqua P., Allred E. N., Moore D. H., Meli S., Gasparini G. Tumor angiogenesis: a new significant and independent prognostic indicator in early-stage breast carcinoma.
J. Natl. Cancer Inst. (Bethesda)
,
84
:
1875
-1887,  
1992
.
11
Fontanini G., Del Mastro L., Bevilacqua G. Determination of angiogenesis in human neoplasms: current prognostic and therapeutic implications.
Forum
,
8
:
128
-142,  
1998
.
12
O’Reilly M. S., Boehm T., Shing Y., Fukai N., Vasios G., Lane W. S., Flynn E., Birkhead J. R., Olsen B. R., Folkman J. Endostatin: an endogenous inhibitor of angiogenesis and tumor growth.
Cell
,
88
:
277
-285,  
1997
.
13
Lin P., Sankar S., Shan S., Dewhirst M. W., Polverini P. J., Quinn T. Q., Peters K. G. Inhibition of tumor growth by targeting tumor endothelium using a soluble vascular endothelial growth factor receptor.
Cell Growth Differ.
,
9
:
49
-58,  
1998
.
14
Brooks P. C., Silletti S., von Schalscha T. L., Friedlander M., Cheresh D. A. Disruption of angiogenesis by PEX, a noncatalytic metalloproteinase fragment with integrin binding activity.
Cell
,
92
:
391
-400,  
1998
.
15
Tozer G. M., Prise V. E., Wilson J., Locke R. J., Vojnovic B., Stratford M. R. L., Dennis M. F., Chaplin D. J. Combretastatin A-4 phosphate as a tumor vascular-targeting agent: early effects in tumors and normal tissues.
Cancer Res.
,
59
:
1626
-1634,  
1999
.
16
Dark G. G., Hill S. A., Prise V. E., Tozer G. M., Pettit G. R., Chaplin D. J. Combretastatin A-4, an agent that displays potent and selective toxicity toward tumor vasculature.
Cancer Res.
,
57
:
1829
-1834,  
1997
.
17
Fleischer A. C. Sonographic depiction of tumor vascularity and flow: from in vivo models to clinical applications.
J. Ultrasound Med.
,
19
:
55
-61,  
2000
.
18
Evans S. M., Laughlin K. M., Pugh C. R., Sehgal C. M., Saunders H. M. Use of power Doppler ultrasound-guided biopsies to locate regions of tumour hypoxia.
Br. J. Cancer
,
76
:
1308
-1314,  
1997
.
19
Asselin-Paturel C., Lassau N., Guinebretiere J-M., Zhang J., Gay F., Bex F., Hallez S., Leclere J., Peronneau P., Mami-Chouaib F., Chouaib S. Transfer of the murine interleukin-12 gene in vivo by a Semliki Forest virus vector induces B16 tumor regression through inhibition of tumor blood vessel formation monitored by Doppler ultrasonography.
Gene Ther.
,
5
:
606
-615,  
1998
.
20
Cheng W. F., Lee C. N., Chu J. S., Chen C. A., Chen T. M., Shau W. Y., Hsieh C. Y., Hsieh F. J. Vascularity index as a novel parameter for the in vivo assessment of angiogenesis in patients with cervical carcinoma.
Cancer (Phila.)
,
85
:
651
-657,  
1999
.
21
Fleischer A. C., Wojcicki W. E., Donnelly E. F., Pickens D. R., Thirsk G., Thurman G. B., Hellerqvist C. G. Quantified color Doppler sonography of tumor vascularity in an animal model.
J. Ultrasound Med.
,
18
:
547
-551,  
1999
.
22
Sehgal C. M., Arger P. H., Rowling S. E., Conant E. F., Reynolds C., Patton J. A. Quantitative vascularity of breast masses by Doppler Imaging: regional variations and diagnostic implications.
J. Ultrasound Med.
,
19
:
427
-440,  
2000
.
23
Eriksson R., Persson H. W., Dymling S. O., Lindstrom K. Evaluation of Doppler ultrasound for blood perfusion measurements.
Ultrasound Med. Biol.
,
17
:
445
-452,  
1991
.
24
Rubin J. M., Bude R. O., Carson P. L., Bree R. L., Adler R. S. Power Doppler US: a potentially useful alternative to mean frequency-based color Doppler US.
Radiology
,
190
:
853
-856,  
1994
.
25
Lencioni R., Pinto G., Armillotta N., Bartolozzi C. Assessment of tumor vascularity in hepatocellular carcinoma: comparison of power Doppler US and color Doppler US.
Radiology
,
201
:
353
-358,  
1996
.
26
Kripke M. Speculations on the role of ultraviolet radiation in the development of malignant melanoma.
J. Natl. Cancer Inst. (Bethesda)
,
63
:
541
-545,  
1979
.
27
Fidler I. J. Biological behavior of malignant melanoma cells correlated to their survival in vivo.
Cancer Res.
,
35
:
218
-224,  
1975
.
28
Murphy G. P., Hrushesky W. J. A murine renal cell carcinoma.
J. Natl. Cancer Inst. (Bethesda)
,
50
:
1013
-1025,  
1973
.
29
Coughlin C. M., Salhany K. E., Gee M. S., LaTemple D. C., Kotenko S., Ma X. J., Gri G., Wysocka M., Kim J. E., Liu L., Liao F., Farber J. M., Pestka S., Trinchieri G., Lee W. M. F. Tumor cell responses to IFN-γ affect tumorigenicity and response to IL-12 therapy and antiangiogenesis.
Immunity
,
9
:
25
-34,  
1998
.
30
Gee M. S., Koch C. J., Evans S. M., Jenkins W. T., Pletcher C. H., Moore J. S., Koblish H. K., Lee J., Lord E. M., Trinchieri G., Lee W. M. F. Hypoxia-mediated apoptosis from angiogenesis inhibition underlies tumor control by recombinant interleukin 12.
Cancer Res.
,
59
:
4882
-4889,  
1999
.
31
Majewski S., Marczak M., Szmurlo A., Jablonska S., Bollag W. Interleukin-12 inhibits angiogenesis induced by human tumor cell lines in vivo.
J. Invest. Dermatol.
,
106
:
1114
-1118,  
1996
.
32
Thurston G., Baluk P., Hirata A., McDonald D. M. Permeability-related changes revealed at endothelial cell borders in inflamed venules by lectin binding.
Am. J. Physiol
,
271
:
H2547
-H2562,  
1996
.
33
Grosios K., Holwill S. E., McGown A. T., Pettit G. R., Bibby M. C. In vivo and in vitro evaluation of combretastatin A-4 and its sodium phosphate prodrug.
Br. J. Cancer
,
81
:
1318
-1327,  
1999
.
34
Goertz D. E., Christopher D. A., Yu J. L., Kerbel R. S., Burns P. N., Foster F. S. High-frequency color flow imaging of the microcirculation.
Ultrasound Med. Biol.
,
26
:
63
-71,  
2000
.
35
Meyerowitz C. B., Fleischer A. C., Pickens D. R., Thurman G. B., Borowsky A. D., Thirsk G., Hellerqvist C. G. Quantification of tumor vascularity and flow with amplitude color Doppler sonography in an experimental model: preliminary results.
J. Ultrasound Med.
,
15
:
827
-833,  
1996
.
36
Taylor J. S., Tofts P. S., Port R., Evelhoch J. L., Knopp M., Reddick W. E., Runge V. M., Mayr N. MR imaging of tumor microcirculation: promise for the new millennium.
J. Magn. Reson. Imaging
,
10
:
903
-907,  
1999
.
37
Bhujwalla Z. M., Artemov D., Glockner J. Tumor angiogenesis, vascularization, and contrast-enhanced magnetic resonance imaging.
Top. Magn. Reson. Imaging
,
10
:
92
-103,  
1999
.