Phase II Study of the Antiangiogenic Agent SU5416 in Patients with Advanced Soft Tissue Sarcomas

5732 Vol. 10, 5732–5740, September 1, 2004 Clinical Cancer Research

John V. Heymach,1 Jayesh Desai,1 The most common grade 3 toxicities were headache and Judith Manola,1,2 Darren W. Davis,8 thrombosis. Other less serious toxicities included fatigue, David J. McConkey,8 David Harmon,4 nausea, and abdominal pain. The median systolic blood 4 1 pressure increased from 118 mmHg at baseline to 133 after Post-treatment tumor .(0.01 ؍ David P. Ryan, Geraldine Goss, 1 month of treatment (P 1 3 Travis Quigley, Annick D. Van den Abbeele, biopsies showed no significant decreases in VEGF receptor 7 5 Stuart G. Silverman, Susan Connors, phosphorylation compared with baseline in 3 evaluable pa- Judah Folkman,5 Christopher D. M. Fletcher,6 tients. One patient with gastrointestinal stromal tumor who and George D. Demetri1 had rapid progression during SU5416 treatment was subse- Departments of 1Medical Oncology, 2Biostatistical Science, and quently treated with another KIT inhibitor, imatinib mesy- 3Nuclear Medicine, Dana-Farber Cancer Institute, Boston, late, and had a partial response lasting >36 months. Massachusetts; 4Division of Hematology-Oncology, Massachusetts Conclusions: SU5416 was relatively well tolerated but 5 General Hospital, Boston, Massachusetts; Department of Surgery, did not demonstrate significant antitumor activity against Children’s Hospital, Boston, Massachusetts; Departments of advanced soft tissue sarcoma. Correlative studies suggest 6Pathology and 7Radiology, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts; and 8The University of that VEGF receptor or KIT inhibition was incomplete in at Texas M. D. Anderson Cancer Center, Houston, Texas least some cases, providing a possible explanation for the observed lack of activity. ABSTRACT Purpose: SU5416 (semaxanib) is a small molecule inhib- INTRODUCTION itor of the vascular endothelial growth factor (VEGF) The growth and metastatic spread of tumors are dependent receptor-2 and KIT receptor tyrosine kinases. This Phase II on angiogenesis, a multistep process regulated by the local study was conducted to investigate the safety and efficacy of balance of endogenous pro- and antiangiogenic factors (1). SU5416 for patients with soft tissue sarcomas. Among the known stimulators of tumor angiogenesis, vascular Experimental Design: Thirteen patients with locally ad- endothelial growth factor (VEGF) is thought to play a pivotal vanced or metastatic soft tissue sarcomas were treated with role. The activities of VEGF are primarily mediated by two SU5416 via intravenous infusion at a dose of 145 mg/m2 receptors, VEGF receptor-1 (VEGFR-1 and Flt-1) and twice weekly. In selected cases tumor biopsies were taken VEGFR-2 (KDR and Flk-1; ref. 2). VEGF-induced endothelial before and after 2 months of treatment. migration, proliferation, and vascular permeability appear to be Results: The median progression-free survival was 1.8 dependent on VEGFR-2 signaling via its tyrosine kinase do- months. Median overall survival was 22.8 months. No ob- main. In preclinical models, blockade of the VEGF pathway jective tumor responses were observed. There was evidence inhibits tumor growth as demonstrated by investigators using of shorter survival among patients with high baseline urine monoclonal antibodies that bind VEGF or VEGFR-2 or small No grade 4 toxicities were observed. molecule inhibitors of the VEGFR-2 tyrosine kinase domain .(0.04 ؍ VEGF levels (P (3–6). These approaches are currently being evaluated in human clinical trials. SU5416 (semaxanib) is a small molecule inhibitor of Received 1/26/04; revised 5/504; accepted 5/11/04. VEGFR-2 tyrosine kinase (3). In murine models, SU5416 re- Grant support: Biostatistics support provided by Grant CA-06516 duced tumor angiogenesis and inhibited the growth of a broad from the National Cancer Institute. J. Heymach is supported in part by range of tumor types (3, 7). Phase I and II clinical trials have American Society for Clinical Oncology Young Investigator Award and been conducted using a twice-weekly dosing schedule, with a American Association for Cancer Research-Amgen Research Fellow- ship. G. Demetri and J. Desai are supported in part by the Richard I. maximum tolerated dose of 145 mg/m2. Dose-limiting toxicities Kaner Ewing’s Sarcoma Research Fund, the Leslie’s Links Foundation, were observed including nausea, vomiting, and headache (8, 9). the Abraham J. and Phyllis Katz Foundation, the Quick Family Sarcoma A once-weekly schedule has also been tested, which also Research Fund, and the Ludwig Trust for Cancer Research at Harvard showed that doses up to 145 mg/m2 were well tolerated (10). Medical School. D. Davis is supported in part by the Commonwealth Evidence of antitumor activity has been observed in several Foundation for Cancer Research. The costs of publication of this article were defrayed in part by the studies (10–12). When SU5416 was given in combination with payment of page charges. This article must therefore be hereby marked gemcitabine and cisplatin, a high incidence of thromboembolic advertisement in accordance with 18 U.S.C. Section 1734 solely to complications was observed and was accompanied by labora- indicate this fact. tory evidence of endothelial cell perturbation, discouraging ad- Requests for reprints: George Demetri, Center for Sarcoma and Bone Oncology, Dana-Farber Cancer Institute, SW530, 44 Binney Street, ditional use of this combination (13). After initiation of this trial, Boston, MA 02115.E-mail: [email protected]. it was also reported that SU5416 also inhibits the KIT receptor ©2004 American Association for Cancer Research. tyrosine kinase (14, 15). Mutations causing constitutive activa-

tion of the KIT receptor appear to play a critical role in the apy Evaluation Program of the National Cancer Institute. The pathogenesis of gastrointestinal stromal tumors and a KIT in- drug was administered via intravenous infusion in an outpatient hibitor, imatinib, is effective therapy for the treatment of gas- setting at a dose of 145 mg/m2 twice weekly. The first dose of trointestinal stromal tumor (16–19). SU5416 was administered at 100 mL/hour for the first 15 Soft tissue sarcomas are a diverse group of tumors of minutes before increasing to full speed (200 mL/hour). If the mesenchymal origin. For most histologic subsets of advanced or patient had any evidence of venous irritation, the infusion rate metastatic soft tissue sarcomas, chemotherapy is currently the was slowed to 100 mL/hour. Patients were observed for a treatment of choice. Unfortunately, these drugs carry substantial minimum of 3 hours after their first three doses. Components in toxicity, and resistance tends to arise quickly. The vast majority the formulation included Cremophor (poloxyl 35 castor oil), of soft tissue sarcomas were found to express VEGF, and polyethylene glycol 400, benzyl alcohol, and dehydrated alcohol elevated serum VEGF levels were found to correlate with higher (supplied by Sugen, Redwood City, CA). To prevent allergic histologic tumor grade (20, 21). VEGF receptor inhibitors reaction to Cremophor, premedication with dexamethasone and showed antitumor activity in preclinical models of sarcoma (3, an H1-receptor blocker were used. Patients also received pro- 7). Therefore, soft tissue sarcomas represent a rational target in phylaxis against thromboembolic disease with coumadin 1 mg/ which to perform therapeutic testing of VEGF pathway inhibi- day, Dalteparin 2,500 IU subcutaneously/daily, or Lovenox 30 tors. For this reason, we have conducted a Phase II study of mg subcutaneously/daily. SU5416 in patients with advanced or metastatic soft tissue Evaluations During Treatment. Pretreatment evalua- sarcoma. We have also explored potential markers of antiangio- tion was carried out within 14 days before treatment and in- genic activity including urine VEGF and basic fibroblast growth cluded informed consent, a complete history, and physical ex- factor (bFGF) levels and tumor endothelial cell VEGFR phos- amination. Baseline laboratory studies included complete blood phorylation assessed by laser scanning cytometry (22). count with differential, comprehensive coagulation profile, liver function tests, renal function, serum chemistries, urinalysis, and pregnancy test. Baseline electrocardiogram; chest X-ray; com- MATERIALS AND METHODS puted tomography scans of the chest, abdomen, and pelvis with Patient Eligibility. All of the patients were required to quantification of measurable disease; and bone scan (if relevant) have histologically documented soft tissue sarcoma with meas- were also performed. 2-Deoxy-2-[18F]fluoro-D-glucose positron urable disease that was metastatic or inoperable. This study was emission tomography (18FDG-PET) scans were also performed designed to include patients who were treated previously with to complete the baseline assessment. During treatment, patients chemotherapy and who had either not responded or had pro- were seen by a physician weekly and had a complete physical gression of disease after an initial response. All of the patients examination, assessment of performance status, complete blood had to have had prior chemotherapy. All of the patients were count, serum chemistries, coagulation profile, and urine and Ն18 years of age, signed a written informed consent, and had an serum assays for angiogenic factors performed. At each visit, Eastern Cooperative Oncology Group performance status of patients were also evaluated for toxicity, which was graded 0–1, with a life expectancy Ն12 weeks. Other inclusion criteria according to the National Cancer Institute Common Toxicity included adequate hematologic function (white blood cell count Criteria. Patients had a 60-minute corticotrophin stimulation test Ͼ2,000 cells/␮L and platelet count Ͼ100,000 cells/␮L), renal performed every 4 weeks while on study to monitor for adrenal function (serum creatinine Յ1.5 mg/dL), hepatic function (se- insufficiency. Disease response was assessed every 8 weeks rum bilirubin Յ the upper limit of normal and transaminases with tumor imaging via computed tomography scans and PET Ͻ1.5 times the upper limit of normal) and coagulation function scans. Response to therapy was evaluated according to Re- (prothrombin and partial thromboplastin time Ͻ1.25 times the sponse Evaluation Criteria in Solid Tumors criteria (23). Where upper limit of normal, fibrin split products Յ1.0 ␮g/mL, and possible, accessible tumor was rebiopsied after 8 weeks of fibrinogen Ͼ200 mg/dL). At least 3 weeks had to have elapsed treatment for evaluation of tumor vasculature. Urine was col- from any prior radiotherapy or chemotherapy (6 weeks from any lected before therapy on days 1, 8, 15, and 22 of each cycle, prior mitomycin C or nitrosurea therapy) and 4 weeks from any centrifuged to remove debris, stored at Ϫ70°C, and batched for major surgery. Exclusion criteria included patients with brain enzyme-linked immunosorbent assay (ELISA) analysis. VEGF metastases, a bleeding diathesis, a history of venous or arterial and bFGF levels were measured using the Quantikine human thrombosis (including pulmonary embolism) in the previous 3 VEGF and Quantikine High Sensitivity Human bFGF ELISA months, a history of allergic reaction to poloxyl 35 castor oil kits (R&D Systems, Inc., Minneapolis, MN) as per the manu- (Cremophor) or paclitaxel, patients with uncompensated cardiac facturer’s instructions. The limits of detection were 5 pg/mL and disease, diabetes mellitus, severe peripheral vascular disease, 0.2 pg/mL for VEGF and bFGF ELISAs, respectively. and patients who were pregnant or breastfeeding. Women of Dose Adjustments. Dose adjustments were based on the childbearing potential were required to have a negative preg- worst toxicity occurring in the previous week of treatment. For nancy test before enrollment, and both partners were required to any grade 3 or 4 hematologic toxicity, drug was withheld until practice suitable contraception during the study. All of the recovery to baseline and was then recommenced at 110 mg/m2. patients were required to sign a written informed consent ap- Patients with Նgrade 3 nausea and vomiting, despite treatment, proved by the Dana-Farber/Partners Cancer Care Institutional had the dose of drug reduced to 110 mg/m2. Patients Review Board before treatment could be started. with Նgrade 2 neurologic toxicity were restarted at 110 mg/m2 Treatment Plan. SU5416 (produced by Sugen, Inc., once recovery to Յgrade 1 was reached. Patients with all South San Francisco, CA) was obtained from the Cancer Ther- other Նgrade 3 nonhematologic toxicities had drug withheld

until recovery to Յgrade 1, with recommencement at 110 mg/ standardized relative to total VEGFR-2. At baseline, phos- m2. In patients with recurrence of any Նgrade 3 toxicity, dose phorylated VEGFR was detected in 30–90% of CD31ϩ was additionally reduced to 85 mg/m2, followed by reduction to endothelial cells. 65 mg/m2. Patients who did not recover from any Նgrade 3 toxicity to Յgrade 1 were removed from study. Patients expe- RESULTS riencing grade 3 or 4 anaphylactic reactions, despite full-dose Patient Characteristics. A total of 15 patients were en- dexamethasone premedication, were removed from study. rolled from May 2000 to October 2000. Two patients were Statistical Considerations. The primary objective of registered to study but never treated with SU5416: 1 patient died this study was to determine the progression-free rate at 6 unexpectedly before receiving any protocol therapy, and the months. The secondary objective was to determine the overall second patient was excluded because of a recent history of deep response rate (partial and complete response). Historically, sur- venous thrombosis. Thus, there were 13 treated patients evalu- vival for patients with metastatic soft tissue sarcomas is very able for this study. Ͻ poor. Long-term disease-free survival is observed in 5% of Table 1 shows patient characteristics at study entry. The ϳ patients, with objective complete response rates being 15%. median age was 49 years (range, 22–72), and the majority (62%) This study was designed to accrue 60 patients in two stages to were rated as Eastern Cooperative Oncology Group perform- distinguish a 15% response rate from a null rate of 5% but ance status 1. The patients were heavily pretreated, with 54% terminated before reaching the first-stage accrual goal due to a having received three or more prior chemotherapy regimens and lack of antitumor activity observed. 85% treated previously with radiotherapy. Synovial sarcoma Descriptive statistics were used to characterize patients at was the most common histologic subtype (31%). baseline and laboratory correlates. The method of Kaplan and Toxicities and Physiologic Changes Associated with Meier was used to estimate overall and progression-free sur- SU5416 Administration. Table 2 summarizes all of the re- vival. We used 95% exact binomial confidence intervals to ported toxicities of grade 2 or higher. Toxicities were rated as estimate the response rate and the proportion progression-free at whether they were attributable to SU5416 treatment, but this 6 months. ANOVA was used to test for a change over time in table shows all of the toxicities regardless of attribution. There blood pressure. were no grade 4 toxicities observed. No patients developed Correlative Studies of Tumor Biopsies. Laser scan- life-threatening or lethal toxicities. The most frequent toxicity ning cytometry was performed using a laser scanning cytom- was headache, which was grade 3 in 4 cases (31%). Headache etry instrument (CompuCyte, Corporation, Cambridge, MA) typically started within several hours after receiving SU5416, consisting of a base unit containing an Olympus BX50 flu- was frequently associated with nausea, and usually resolved orescent microscope and an optics unit coupled to an Argon within 24 hours. Grade 2 rash occurred in 2 patients (15%) and and HeNe laser. The laser scanning cytometry analysis and erythema multiforme in 1 patient (8%), which could be attrib- identification of endothelial cells was performed as described previously (22, 24). Briefly, endothelial cells were stained using a monoclonal antihuman CD31 (clone JC/70A, Dako Corporation, Carpinteria, CA) followed by Cy5-conjugated Table 1 Patient characteristics goat antimouse secondary (Jackson ImmunoResearch Labo- No. (%) ratories, West Grove, PA). For detection of total VEGFR-2, Age formalin-fixed tissues were incubated with rabbit anti-Flk-1 Median 49 antibody (C-20, Santa Cruz Biotechnology, Santa Cruz, CA) Range 22–72 followed by a secondary goat antirabbit IgG conjugated to Sex Male 7 (54) Cy5 (Jackson ImmunoResearch Laboratories). For detection Female 6 (46) Primary histology Synovial sarcoma 4 (31) of phospho-VEGF, a rabbit antiphospho-VEGFR-2 antibody Leiomyosarcoma 3 (23) (Ab-1, Oncogene, La Jolla, CA) was used followed by bioti- Angiosarcoma 2 (15) nylated goat antirabbit IgG (Biocare Medical, Walnut Creek, Fibrosarcoma 1 (8) CA) and Avidin- fluorescein isothiocyanate (PharMingen, Gastrointestinal stromal tumor 1 (8) Desmoplastic small round cell tumor 1 (8) San Diego, CA). Cells were counterstained with propidium Sarcoma, nitric oxide synthase 1 (8) iodide (1 ␮g/mL) for identification of cell nuclei. The rela- Tumor grade Intermediate 1 (10) tive levels of fluorescence for each cell were plotted on a High 9 (90) scattergram. The analytical gates define four quadrants that Not graded/unknown 3 determine the total number of cells within each population Performance status 0 5 (38) Ϫ Ϫ ϩ ϩ 1 8 (62) (i.e., CD31 /pVEGFR , CD31 /pVEGFR , and so forth). Number of prior None 1 (8) Staining with negative controls was performed in parallel as surgeries 1 4 (31) controls and were used to set analytical gates. Each slide was 2–4 5 (39) Ͼ placed on a computer-controlled motorized stage, and the 4 (5, 6, or 8) 3 (23) Prior radiotherapy No 2 (15) area to be analyzed was visually identified using the epifluo- Yes 11 (85) rescent microscope of the instrument, excluding normal tis- Prior chemotherapy 1 2 (15) sue and necrotic regions. The sections were then subjected to regimens 2 4 (31) quantitative analysis. Three to five randomly selected areas 3 5 (39) were quantitatively assessed on each slide, and levels were 4 or more (4, 6) 2 (15)

Table 2 National Cancer Institute Common Toxicity Criteria grade Hypertension has been reported in clinical trials for several 2 and 3 toxicities VEGF pathway antagonists (28, 29). No patients in this study Grade experienced grade 2 or higher hypertension. There was, however, a statistically significant increase in the median systolic 2 3 Adverse event (N ϭ 13 total patients) N N blood pressure from 118 mmHg at baseline to 133 mmHg after 1 month of treatment (P ϭ 0.01; Table 3). This difference had Headache 2 4 Fatigue 4 - disappeared by the end of cycle 2. The diastolic blood pressure Dyspnea 2 1 decreased slightly during treatment, whereas the pulse pressure Abdominal pain 3 - (systolic minus diastolic blood pressure) increased from base- Thrombosis/embolism - 2 line to cycle 1 (P ϭ 0.007). Nausea 1 1 Salivary gland changes 2 - SU5416 was also reported recently to cause secondary Rash 2 - polycythemia vera in 3 patients with Von Hippel-Lindau disease Anemia 2 - (30). No significant change in hematocrit was observed during Hyperglycemia 2 - treatment in this study (data not shown), suggesting that Insomnia 2 - Vomiting - 1 this effect may be specific to patients with von Hippel-Lindau Elevated partial thromboplastin time - 1 disease. Supraventricular arrhythmias - 1 Treatment Duration. The median duration of treatment Bone pain - 1 was two cycles. One patient with leiomyosarcoma received four Chest pain - 1 cycles before progressing. Treatment was discontinued in 12 Leukopenia 1 - Fever 1 - patients for disease progression and in 1 patient for toxicity Erythema multiforme 1 - (deep vein thrombosis as discussed above). Pruritis 1 - Response. Using formal Response Evaluation Criteria in Anorexia 1 - Solid Tumors criteria to assess response to treatment, 3 of 13 Myalgia 1 - Pelvic pain 1 - treated patients were inevaluable for response. Two of these Pleuritic pain 1 - patients demonstrated clinical evidence of progression, whereas Cough 1 - the third patient was removed from treatment due to toxicity before completing two cycles but had no evidence of progression. One patient had a follow-up scan before completion of two cycles. At the time of the scan, tumor measurements indicated utable to SU5416 or the Cremophor vehicle used in the drug stable disease. This patient demonstrated clinical progression by formulation. the end of cycle 2 though, and is, thus, considered to have Patients received prophylaxis against thromboembolic dis- progressed. There were no responses observed in 13 evaluable ease with either coumadin or low molecular weight heparin patients (0%, 95% one-sided exact binomial confidence interval, during the study because of an increased incidence of thrombo- 0–21%). embolic events, particularly catheter-associated clots, noted in Fig. 1A shows the progression-free survival. The median the Phase I studies (13, 25–27). One patient developed an time to progression was 1.8 months. The study was originally internal jugular vein thrombosis on day 10 of the first treatment designed to estimate the proportion of patients free of progres- cycle in the context of having an internal jugular Port-a-Cath in sion at 6 months, because that was considered to be potentially place. This occurred despite prophylactic anticoagulation with reflective of clinical activity and stabilization of disease. We low-dose warfarin. Study drug was withheld, and the patient observed a 3-month progression-free rate of 8% (95% confi- was fully anticoagulated with enoxaparin. The port was left in dence interval, 1–54%) and the 6-month progression-free rate of place after radiologic studies revealed it to be fully patent. 0% (95% confidence interval, 0–22%). SU5416 was recommenced on resolution of symptoms, 3 weeks The overall survival is shown in Fig. 1B. At the time of this later. Ten days after recommencing SU5416, despite being fully analysis, 7 patients have died and 6 patients remain alive. anticoagulated, he went on to develop a recurrence of the Median overall survival is 22.8 months. thrombosis. Study drug was promptly stopped, and the patient One informative case study requires expansion. A 57-year- was removed from protocol. There were no other serious hema- old man diagnosed in 1998 with metastatic gastrointestinal tologic toxicities. Two patients (15%) experienced grade 2 ane- stromal tumor was treated previously with four chemotherapy mia, and 1 patient (8%) developed grade 2 leukopenia. regimens and began treatment with SU5416 on August 1, 2000.

Table 3 Median blood pressure (mmHg) and pulse pressure (systolic minus diastolic pressure) for patients treated with SU5416 Systolic (range) Diastolic (range) Pulse pressure (range) Baseline 118 (86–142) 78 (51–92) 44 (30–50) Cycle 1 133 (94–165)* 77 (57–94) 55 (32–77)† Cycle 2 116 (104–142) 74 (48–86)* 47 (38–75) * Rank sum P value compared with baseline ϭ 0.01. † Rank sum P value compared with baseline ϭ 0.007.

time to progression among patients with high (greater than median) baseline VEGF levels, but power is limited by a small sample size (Fig. 3A). As seen in Fig. 3B, there was evidence of shorter survival time among patients with high baseline VEGF levels (P ϭ 0.04). There was no difference in survival or time to progression associated with baseline bFGF levels (data not shown). Tumor Biopsies. To directly assess the effects of SU5416 on the tumor, we used laser scanning cytometry to measure VEGFR phosphorylation status from biopsies taken at

Fig. 1 Kaplan-Meier curves for progression-free survival (PFS; A) and overall survival (OS; B). Median PFS was 1.8 months, and the median OS was at the time of this analysis was 22.8 months. Dashed lines, 95% confidence interval.

After two cycles of therapy, he had evidence of disease progression by computed tomography and 18FDG-PET scans (Fig. 2, A and B) and discontinued SU5416 in early October, 2000. In November 2000 he initiated therapy with imatinib mesylate (Gleevec, STI571), and by the end of the first cycle a dramatic Fig. 2 18FDG-PET of patient with gastrointestinal stromal tumors reduction in tumor FDG uptake was observed (Fig. 2, C and D), treated initially with SU5416 who subsequently had a durable partial which was accompanied by tumor shrinkage by computed to- response to imatinib. A, baseline scan before initiating therapy. Arrow ϳ mography. The patient has had a partial response to imatinib, shows location of large hepatic metastases. B, after 2 months of Ͼ treatment with SU5416. Patient was discontinued from protocol for which has lasted 3 years at the time of this analysis. tumor progression as judged by computed tomography scan. C, scan 1 18 FDG-PET Scans. To assess changes in tumor glucose month later, before initiating therapy with imatinib. D, dramatic reduc- uptake and metabolism during therapy, 18FDG-PET scans were tion in FDG-avid disease after 1 month of therapy with imatinib. Patient performed. In 7 of 7 patients assessed at baseline and at the end has been treated with imatinib for Ͼ3 years at the time of this analysis of cycle 2, an increase in tumor 18FDG uptake was observed. without evidence of progression. Computed tomography scans performed for all 7 of the patients at the end of cycle 2 showed concordant tumor progression. Angiogenic Markers. Table 4 shows urinary bFGF and Table 4 VEGF and bFGF levels (pg/ml) at baseline and end of VEGF levels for patients at baseline and after cycle 1 of treat- cycle 1 ment. The median urinary VEGF and bFGF levels increased by Baseline Cycle 1 Change (%) 21% and 14%, respectively over this interval. For 11 patients VEGF Median 101 122 ϩ21 with both evaluable progression status at 6 months and known Range 31–605 37–1,000 laboratory levels, differences in time to progression by baseline basic FGF Median 1.96 2.23 ϩ14 VEGF levels were examined. There was a trend toward shorter Range 5.06–11.41 1.11–40.38

baseline and after two cycles of therapy. Of the 13 patients known to be a regulator of vascular tone and permeability, in treated in the study, 5 underwent repeat biopsies, yielding three part through modulation of endothelial nitric oxide synthase (37, pairs of evaluable biopsies (in 1 case, the second biopsy yielded 38). By inhibiting VEGFR-2, SU5416 may interfere with these necrotic tissue that could not be evaluated; in another, the processes and inhibit nitric oxide-mediated vasodilation, caus- amount of tissue was not adequate for a full analysis). Treatment ing a change in vessel compliance. with SU5416 did not lead to significant decreases in VEGFR On the basis of preclinical studies, VEGF pathway inhib- phosphorylation in any of these three pairs (Table 5). In 2 cases, itors, if active, would be expected to result in disease stabiliza- VEGFR phosphorylation decreased slightly after treatment, whereas in 1 case it increased. Interestingly, 1 of the cases in which partial inhibition was observed was the aforementioned patient with gastrointestinal stromal tumor who subsequently had a prolonged clinical response to imatinib, providing additional evidence for incomplete target inhibition by the drug. An analysis of additional angiogenic markers using tumor biopsies from this study and two other clinical trials using SU5416 or a related molecule, SU6668, is detailed in a separate study.9

DISCUSSION For most advanced or metastatic soft tissue sarcoma, conventional cytotoxic chemotherapy remains standard for pallia- tive management of patients. Unfortunately these tumors are relatively chemoresistant, and responses to these chemotherapy agents are usually suboptimal and short-lived. The vast majority of soft tissue sarcomas express VEGF (20, 21). Preclinical studies indicate that VEGF pathway blockers inhibit tumor growth in animal models of sarcoma and other malignancies (4, 7, 31), suggesting that VEGFR-2 may be a viable target. To test this hypothesis, we conducted a Phase II trial of the small molecule VEGFR-2 inhibitor SU5416. In Phase I testing, 145 mg/m2 twice weekly was deter- mined to be a dose that was acceptable for chronic dosing and resulted in a systemic exposure comparable with what was required in animals for effective tumor inhibition (8, 32). This dose has subsequently been used for Phase II testing (12, 25–27, 33–35). At higher doses, dose-limiting toxicities were observed including headache, nausea, and vomiting. We found that SU5416 could be safely administered chronically at this dose. There were no grade 4 toxicities observed. Consistent with previous studies, headache was the most frequent grade 3 toxicity in this study and was typically reversible within 24 h. Certain toxicities, such as hyperglycemia and insomnia, were Fig. 3 Relationship between baseline VEGF levels and clinical out- likely attributable to the steroid premedication. come. A, progression-free survival and (B) overall survival in patients It is interesting to note that systolic blood pressure and with high (Նmedian level of 101 pg/ml) or low (Ͻ101 pg/ml) baseline pulse pressure increased slightly after one cycle, although this levels of urinary VEGF. A shorter overall survival time was observed difference had disappeared by the end of cycle 2 (Table 3). for patients with high baseline VEGF levels (P ϭ 0.04). Hypertension has been reported for bevacizumab (29, 36) and VEGFR-2 antagonists, raising the possibility that this effect may be related to blockade of the VEGF/VEGFR-2 pathway. The Table 5 Quantitative analysis of VEGFR-2 phosphorylation by laser mechanism by which VEGFR-2 blockade might cause an in- scanning cytometry crease in blood pressure has not been established. VEGF is Phospho-VEGFR Histology Pre (%) Post (%) PFS (days) Angiosarcoma 32.2 Ϯ 4.1 26.5 Ϯ 5.7 35 GIST 45.9 Ϯ 0.4 29.6 Ϯ 14.3 65 Leiomyosarcoma 33.4 Ϯ 3.0 45.2 Ϯ 1.3 66 9 Darren W. Davis, Henry Q. Xiong, Roy S. Herbst, Walter M. Stadler, John V. Heymach, George D. Demetri, Asif Rashid, Yu Shen, Sijen NOTE. Paired tumor biopsies from 3 patients pre- and post- Wen, Steven Canfield, James L. Abbruzzese, and David J. McConkey. treatment were assessed. Values represent mean percent of phospho- Pharmacodynamic analysis of target inhibition and tumor cell death in VEGFRϩ endothelial cells Ϯ SE. biopsies obtained from patients treated with the VEGF receptor antag- Abbreviations: PFS, progression-free survival; GIST, gastrointes- onists SU-5416 or SU-6668. Submitted for publication. tinal stromal tumor.

tion as opposed to tumor shrinkage. Conventional radiologic VEGFR phosphorylation decreased slightly after therapy, means of assessing activity based on response rate may, there- whereas in the third patient, an increase was observed (Table 5). fore, not be an optimal indicator of the biological activity of Because of the small number of paired specimens analyzed and these agents. In view of this, it has been proposed that progres- the absence of clinical responses to SU5416, it is not possible to sion-free survival or progression-free rate may be a more ap- determine whether the degree of inhibition correlates with clin- propriate end point for Phase II studies of agents expected to be ical response. For comparison, however, in a murine orthotopic cytostatic (39, 40). bladder tumor model, a monoclonal antibody inhibitor of murine In this study we observed no objective tumor responses out VEGFR-2 effectively inhibited tumor growth and led to a 50% of 13 evaluable patients. The median progression-free survival inhibition of VEGFR phosphorylation as assessed by Laser was 1.8 months, whereas the 3- and 6-month progression-free scanning cytometry.9 These results are consistent with the hy- rates were 8% (95% confidence interval,1–54%) and 0% (95% pothesis that SU5416 incompletely inhibited its intended targets confidence interval, 0–22%), respectively. The results were not in this study. Similar findings were observed in a set of 8 considered sufficiently promising to continue accrual to study, additional patients treated with SU5416 or SU6668 in other and the production and testing of SU5416 was ceased by the studies.9 biopharmaceutical firm that manufactured it. Due to the small Incomplete receptor inhibition may be a result of inade- sample size, the latter confidence interval includes both 5% (the quate systemic exposure. In Phase I testing an elimination rate signifying minimal activity) and 15% (the rate indicating half-life of Ͻ1 hour was observed (8). Furthermore, using the SU5416 was worthy of additional study). Given this broad twice weekly schedule, a 50–60% induction of SU5416 clear- confidence interval, we cannot exclude the possibility that ance was observed with chronic dosing (10, 32, 42) resulting in SU5416 might be active in slowing tumor progression in some lower systemic exposures. A schedule using a 5-day loading cases. The available data suggests, however, that the magnitude dose followed by once a week dosing at 145 mg/m2 prevented of any such activity would be small. In a review of European this induction, leading to consistently higher serum levels Organization for Research and Treatment of Cancer studies of than twice weekly dosing (10). Given the short half-life, dose- patients with advanced soft tissue sarcoma, 380 patients treated limiting toxicities, and induction of SU5416 clearance, sus- previously with chemotherapy had a 3-month progression-free tained levels needed for therapeutic efficacy may not have been rate of 28% and a 6-month progression-free rate of 10% (39). achieved or sustained using a twice-weekly dosing schedule. The subgroup of patients treated with inactive agents had The lack of antitumor response in a patient with gastrointestinal 3-month and 6-month progression-free rates of 21% and 8%, stromal tumor, who subsequently responded to imatinib, addi- respectively. On the basis of these studies, the authors suggest tionally supports this possibility (discussed below). Therefore, that a 3-month progression-free rate of Յ20% suggests inactiv- based on this study, we cannot rule out the possibility that agents ity, whereas a progression-free rate Ն40% suggests activity. achieving a higher degree or more prolonged VEGFR inhibition The 3-month progression-free rate of 8% observed in our study may be useful for the treatment of soft tissue sarcoma. and the absence of objective tumor responses suggest that the It is noteworthy that a patient with gastrointestinal stromal agent does not have significant activity in advanced soft tissue tumor who had progressive disease after 2 months of treatment sarcoma. In other studies of SU5416 a limited degree of anti- with SU5416 subsequently has had a partial response lasting Ͼ3 tumor activity has been observed. In a recently published study, years to imatinib mesylate. Imatinib is thought to exert its which included 26 patients with soft tissue sarcoma treated with antitumor effect through inhibition of the mutated KIT receptor SU5416 at the same dose and schedule as this study, 1 patient kinase, which is constitutively active in most cases of gastroin- had a partial response (4%) and 5 patients had stable disease for testinal stromal tumor (16–18, 43, 44). Interestingly, SU5416 Ն3 months for a 3-month progression-free rate of 23% (12). was reported to inhibit KIT receptor tyrosine phosphorylation ␮ Evidence of antitumor activity has also been observed in pa- induced by stem cell factor with an IC50 of 0.01 mol/L in H526 tients with other malignancies treated with SU5416 (10, 11, small cell carcinoma cells (15), and 0.1–1.0 ␮mol/L in M07e ␮ 33, 34). cells (14), similar to the IC50 of 0.1 mol/L observed for Although SU5416 did not demonstrate significant antitu- imatinib in M07e cells (46). Therefore, SU5416 showed no mor activity in this study, the median survival of 22.8 months antitumor activity in this patient despite its ability to inhibit KIT was longer than would be expected for patients with metastatic tyrosine kinase activity at concentrations similar to or lower soft-tissue sarcoma, which was ϳ12 months in a large analysis than imatinib (14, 15, 44, 45, 46). The dramatic and sustained of patients treated in first-line chemotherapy studies (41). This antitumor activity observed for imatinib, but not SU5416, may is likely explained by the heterogeneous population group in this be due to differences in the dosing schedule (daily for imatinib small study and, in particular, the large percentage of patients versus twice weekly for SU5416), drug pharmacokinetics (the with synovial cell sarcoma (31%), a tumor subtype generally elimination half-life of SU5416 is Ͻ1 hour versus 13–16 hours associated with longer survival (41). for imatinib; Ref. 47), or a difference in sensitivity of the There are a number of possible explanations for the lack of specific activating KIT mutation to either drug. antitumor activity observed. One possibility is that the inhibition Another possible explanation for the overall observed lack of of the VEGFR-2 by SU5416 may not have been complete or antitumor activity is that for advanced tumors with established consistent. To assess the interaction between SU5416 and its vasculature, targeting one receptor for a single angiogenic factor targets, biopsies were performed on a subset of patients at may not be sufficient. Sarcomas are known to produce a variety of baseline and after 2 months of therapy and assessed for VEGFR proangiogenic factors in addition to VEGF, including bFGF, acidic phosphorylation by laser scanning cytometry. In 2 patients fibroblast growth factor, VEGF-B, VEGF-C, interleukin 8, trans-

forming growth factor ␣, and matrix metalloproteinases (20, 48– 2. Matsumoto T, Claesson-Welsh L. VEGF receptor signal transduc- 51), which could presumably help sustain tumor vasculature even tion. Sci STKE 2001;112:1–17. in the presence of VEGFR-2 inhibition. VEGFR-1 and other VEG- 3. Fong TA, Shawver LK, Sun L, et al. SU5416 is a potent and selective FRs may play an important role in tumor vasculature as well (52). inhibitor of the vascular endothelial growth factor receptor (Flk-1/KDR) that inhibits tyrosine kinase catalysis, tumor vascularization, and growth For this reason, combinations of agents or inhibitors with broader of multiple tumor types. Cancer Res 1999;59:99–106. specificity may prove to be more effective. 4. Borgstrom P, Hillan KJ, Sriramarao P, Ferrara N. Complete inhibi- Elevated baseline levels of urinary VEGF but not bFGF tion of angiogenesis and growth of microtumors by anti-vascular endo- appeared to be associated with shorter survival (Fig. 3B), al- thelial growth factor neutralizing antibody: novel concepts of angiostatic though this finding was of borderline significance (P ϭ 0.04). therapy from intravital videomicroscopy. Cancer Res 1996;56:4032–9. Furthermore, there was a trend toward shorter time to progres- 5. Borgstrom P, Bourdon MA, Hillan KJ, Sriramarao P, Ferrara N. sion in patients with high baseline VEGF (Fig. 3A). Interest- Neutralizing anti-vascular endothelial growth factor antibody com- ingly, in a Phase I study of SU5416, 4 patients who obtained pletely inhibits angiogenesis and growth of human prostate carcinoma micro tumors in vivo. Prostate 1998;35:1–10. clinical benefit (defined by tumor regression or diseases stabi- 6. Prewett M, Huber J, Li Y, et al. Antivascular endothelial growth lization for at least 12 weeks) had higher baseline 24-hour factor receptor (fetal liver kinase 1) monoclonal antibody inhibits tumor urinary levels of VEGF than the 18 patients who did not have angiogenesis and growth of several mouse and human tumors. Cancer clinical benefit (323.4 versus 182.7, P Ͻ 0.05; Ref. 10). Because Res 1999;59:5209–18. no patients in this study obtained clinical benefit by these 7. Angelov L, Salhia B, Roncari L, McMahon G, Guha A. Inhibition of criteria, such a comparison is not possible in our study. Never- angiogenesis by blocking activation of the vascular endothelial growth theless, the trend toward shorter time to progression in patients factor receptor 2 leads to decreased growth of neurogenic sarcomas. with high baseline VEGF in this study suggests that elevated Cancer Res 1999;59:5536–41. baseline VEGF would most likely not be useful as a marker of 8. Rosen L, Mulay M, Mayers A, et al. Phase I dose-escalating trial of SU5416, a novel angiogenesis inhibitor in patients with advanced ma- responsiveness to a VEGFR-2 antagonist. lignancies. In: Proc Am Soc Clin Oncol 1999;18:A618. The mean urinary VEGF and bFGF levels were 21% and 14% 9. Stopeck A. Results of a Phase I Dose-escalating study of the anti- higher, respectively, at the end of the first cycle of therapy as angiogenic agent, SU5416, in patients with advanced malignancies. compared with baseline. There was considerable variability in these Proc Am Soc Clin Oncol 2000;19:A802. levels for a given patient during therapy, and no clear pattern of 10. Stopeck A, Sheldon M, Vahedian M, Cropp G, Gosalia R, Hannah change could be discerned. Furthermore, it is not known whether A. Results of a Phase I dose-escalating study of the antiangiogenic changes observed were due to treatment with SU5416, disease agent, SU5416, in patients with advanced malignancies. Clin Cancer Res 2002;8:2798–805. progression, or some other cause. Our findings suggest that meas- 11. Mesters RM, Padro T, Bieker R, et al. Stable remission after urement of levels of angiogenic factors may provide prognostic administration of the receptor tyrosine kinase inhibitor SU5416 in a information. This is consistent with earlier studies in sarcoma and patient with refractory acute myeloid leukemia. Blood 2001;98:241–3. other diseases in which elevated VEGF correlated with higher stage 12. Kuenen BC, Tabernero J, Baselga J, et al. Efficacy and toxicity of and/or poor outcome (20, 51), although in one study of synovial the angiogenesis inhibitor SU5416 as a single agent in patients with sarcoma a correlation was not observed (53). It remains to be seen advanced renal cell carcinoma, melanoma, and soft tissue sarcoma. Clin whether urinary VEGF levels will useful for predicting or moni- Cancer Res 2003;9:1648–55. toring responses to therapy (22). 13. Kuenen BC, Rosen L, Smit EF, et al. Dose-finding and pharmaco- In summary, SU5416, an inhibitor of VEGFR-2 and KIT kinetic study of cisplatin, gemcitabine, and SU5416 in patients with solid tumors. J Clin Oncol 2002;20:1657–67. tyrosine kinases, was relatively well tolerated but did not appear 14. Smolich BD, Yuen HA, West KA, Giles FJ, Albitar M, Cherrington to have significant antitumor activity in patients with soft tissue JM. The antiangiogenic protein kinase inhibitors SU5416 and SU6668 sarcoma. In a limited analysis of 3 patients, significant decreases inhibit the SCF receptor (c-kit) in a human myeloid leukemia cell line in VEGFR photophosphorylation were not observed, suggesting and in acute myeloid leukemia blasts. Blood 2001;97:1413–21. that target inhibition was incomplete. This possibility was ad- 15. Krystal GW, Honsawek S, Kiewlich D, et al. Indolinone tyrosine ditionally supported by a patient with gastrointestinal stromal kinase inhibitors block Kit activation and growth of small cell lung tumor who had tumor progression after two cycles of SU5416 cancer cells. Cancer Res 2001;61:3660–8. but subsequently had a prolonged response to imatinib, another 16. Joensuu H, Roberts PJ, Sarlomo-Rikala M, et al. Effect of the KIT inhibitor. This case highlights the importance of not only tyrosine kinase inhibitor STI571 in a patient with a metastatic gastrointestinal stromal tumor. N Engl J Med 2001;344:1052–6. identifying the critical molecular target(s) for a given tumor 17. Demetri GD, von Mehren M, Blanke CD, et al. Efficacy and safety type, but also determining the duration and degree of target of imatinib mesylate in advanced gastrointestinal stromal tumors. inhibition necessary for antitumor activity. N Engl J Med 2002;347:472–80. 18. Lux ML, Rubin BP, Biase TL, et al. KIT extracellular and kinase domain mutations in gastrointestinal stromal tumors. Am J Pathol 2000; ACKNOWLEDGMENTS 156:791–5. The authors would like to thank Amy Potter for her assistance in 19. Hirota S, Isozaki K, Moriyama Y, et al. Gain-of-function mutations of tabulating clinical trial data and Beth Staron for data management in the c-kit in human gastrointestinal stromal tumors. Science 1998;279:577–80. DFCI Quality Assurance Center for Clinical Trials. 20. Graeven U, Andre N, Achilles E, Zornig C, Schmiegel W. Serum levels of vascular endothelial growth factor and basic fibroblast growth factor in patients with soft-tissue sarcoma. J Cancer Res Clin Oncol REFERENCES 1999;125:577–81. 1. Hanahan D, Folkman J. Patterns and emerging mechanisms of the 21. Kuhnen C, Lehnhardt M, Tolnay E, Muehlberger T, Vogt PM, angiogenic switch during tumorigenesis. Cell 1996;86:353–64. Muller KM. Patterns of expression and secretion of vascular endothelial

growth factor in malignant soft-tissue tumours. J Cancer Res Clin Oncol 37. Fukumura D, Gohongi T, Kadambi A, et al. Predominant role of 2000;126:219–25. endothelial nitric oxide synthase in vascular endothelial growth factor- 22. Davis DW, McConkey DJ, Abbruzzese JL, Herbst RS. Surrogate induced angiogenesis and vascular permeability. Proc Natl Acad Sci markers in antiangiogenesis clinical trials. Br J Cancer 2003;89:8–14. USA 2001;98:2604–9. 23. Therasse P, Arbuck SG, Eisenhauer EA, et al. New guidelines to 38. Kou R, Greif D, Michel T. Dephosphorylation of endothelial nitric- evaluate the response to treatment in solid tumors. European Organiza- oxide synthase by vascular endothelial growth factor. Implications for the tion for Research and Treatment of Cancer, National Cancer Institute of vascular responses to cyclosporin A. J Biol Chem 2002;277:29669–73. the United States, National Cancer Institute of Canada. J Natl Cancer 39. Van Glabbeke M, Verweij J, Judson I, Nielsen OS. Progression-free Inst 2000;92:205–16. rate as the principal end-point for phase II trials in soft-tissue sarcomas. 24. Herbst RS, Mullani NA, Davis DW, et al. Development of biologic Eur J Cancer 2002;38:543–9. markers of response and assessment of antiangiogenic activity in a 40. Korn EL, Arbuck SG, Pluda JM, Simon R, Kaplan RS, Christian clinical trial of human recombinant endostatin. J Clin Oncol 2002;20: MC. Clinical trial designs for cytostatic agents: are new approaches 3804–14. needed? J Clin Oncol 2001;19:265–72. 25. Aklilu M, Kindler HL, Gajewski TF, et al. Toxicities of the anti- 41. Van Glabbeke M, van Oosterom AT, Oosterhuis JW, et al. Prog- angiogenic agent SU5416 in phase II studies. Proc Am Soc Clin Oncol nostic factors for the outcome of chemotherapy in advanced soft tissue 2002;21:1921. sarcoma: an analysis of 2,185 patients treated with anthracycline-con- 26. Kindler HL, Vogelzang NJ, Chien K et al. SU5416 in malignant taining first-line regimens–a European Organization for Research and mesothelioma: a University of Chicago Phase II Consortium Study. Proc Treatment of Cancer Soft Tissue and Bone Sarcoma Group Study. J Clin Am Soc Clin Oncol 2001;20:A1359. Oncol 1999;17:150–7. 27. Eng K, Kindler HL, Stadler WM et al. SU5416 in advanced colo- 42. O’Donnell A, Trigo J, Banerji U, et al. A Phase I trial of the VEGF rectal cancer (CRC): a University of Chicago Phase II Consortium inhibitor SU5416, incorporating dynamic contrast MRI assessment of Study. Proc Am Soc Clin Oncol 2001;20:2215. vascular permeability. Proc Am Soc Clin Oncol 2000;19:A685. 28. Yang JC, Haworth L, Steinberg SM, Rosenberg SA, Novotny WA 43. Rubin BP, Singer S, Tsao C, et al. KIT activation is a ubiquitous randomized double-blind placebo-controlled trial of bevacizumab (anti- feature of gastrointestinal stromal tumors. Cancer Res 2001;61:8118–21. VEGF antibody) demonstrating a prolongation in time to progression in 44. Tuveson DA, Willis NA, Jacks T, et al. STI571 inactivation of the patients with metastatic renal cancer. Proc Am Soc Clin Oncol 2002; gastrointestinal stromal tumor c-KIT oncoprotein: biological and clini- 21:A15. cal implications. Oncogene 2001;20:5054–8. 29. Sledge G, Miller K, Novotny W et al. A Phase II trial of single- 45. Heinrich MC, Griffith DJ, Druker BJ, Wait CL, Ott KA, Zigler agent rhumab VEGF (recombinant humanized monoclonal antibody to AJ. Inhibition of c-kit receptor tyrosine kinase activity by STI 571, a vascular endothelial cell growth factor) in patients with relapsed meta- selective tyrosine kinase inhibitor. Blood 2000;96:925–32. static breast cancer. Proc Am Soc Clin Oncol 2000;19:A5C. 46. Buchdunger E, Cioffi CL, Law N, et al. Abl protein-tyrosine kinase 30. Richard S, Croisille L, Yvart J, et al. Paradoxical secondary poly- inhibitor STI571 inhibits in vitro signal transduction mediated by c-kit cythemia in von Hippel-Lindau patients treated with anti-vascular en- and platelet-derived growth factor receptors. J Pharmacol Exp Ther dothelial growth factor receptor therapy. Blood 2002;99:3851–3. 2000;295:139–45. 31. Brekken RA, Overholser JP, Stastny VA, Waltenberger J, Minna 47. Druker BJ, Talpaz M, Resta DJ, et al. Efficacy and safety of a JD, Thorpe PE. Selective inhibition of vascular endothelial growth specific inhibitor of the BCR-ABL tyrosine kinase in chronic myeloid factor (VEGF) receptor 2 (KDR/Flk-1) activity by a monoclonal anti- leukemia. N Engl J Med 2001;344:1031–7. VEGF antibody blocks tumor growth in mice. Cancer Res 2000;60: 48. Tomlinson J, Barsky SH, Nelson S, et al. Different patterns of 5117–24. angiogenesis in sarcomas and carcinomas. Clin Cancer Res 1999;5: 32. Cropp G, Rosen L, Mulay M, et al. Pharmacokinetics and pharma- 3516–22. codynamics of SU5416 in a Phase I, dose escalating trial in patients with 49. Salven P, Lymboussaki A, Heikkila P, et al. Vascular endothelial advanced malignancies. Proc Am Soc Clin Oncol 1999;18:A619. growth factors VEGF-B and VEGF-C are expressed in human tumors. 33. Fiedler W, Tinnefeld H, Mende T, et al. A Phase II study with SU5416 Am J Pathol 1998;153:103–8. in patients with C-Kit positive AML. Proc Am Soc Clin Oncol 2001;20: 50. Heymach JV. Angiogenesis and antiangiogenic approaches to sar- A1148. comas. Curr Opin Oncol 2001;13:261–9. 34. Zahalsky AJ, Wong RJ, Lis E, et al. Phase II trial of SU5416 in 51. Rutkowski P, Kaminska J, Kowalska M, Ruka W, Steffen J. Cytokine patients with advanced incurable head and neck cancer. Proc Am Soc serum levels in soft tissue sarcoma patients: correlations with clinico- Clin Oncol 2002;21:A902. pathological features and prognosis. Int J Cancer 2002;100:463–71. 35. Gajewski TF, Flickinger S, Heimann R, et al. Phase II study of the 52. Luttun A, Tjwa M, Moons L, et al. Revascularization of ischemic anti-angiogenic Flk-1 tyrosine kinase inhibitor SU5416 in melanoma: tissues by PlGF treatment, and inhibition of tumor angiogenesis, arthritis initial results. Proc Am Soc Clin Oncol 2001;20:A1436. and atherosclerosis by anti-Flt1. Nat Med 2002;8:831–40. 36. Yang JC, Haworth L, Sherry RM, et al. A randomized trial of 53. Kawauchi S, Fukuda T, Tsuneyoshi M. Angiogenesis does not bevacizumab, an anti-vascular endothelial growth factor antibody, for correlate with prognosis or expression of vascular endothelial growth metastatic renal cancer. N Engl J Med 2003;349:427–34. factor in synovial sarcomas. Oncol Rep 1999;6:959–64.

Downloaded from clincancerres.aacrjournals.org on September 30, 2021. © 2004 American Association for Cancer Research. Phase II Study of the Antiangiogenic Agent SU5416 in Patients with Advanced Soft Tissue Sarcomas

John V. Heymach, Jayesh Desai, Judith Manola, et al.

Clin Cancer Res 2004;10:5732-5740.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/10/17/5732

Cited articles This article cites 51 articles, 22 of which you can access for free at: http://clincancerres.aacrjournals.org/content/10/17/5732.full#ref-list-1

Citing articles This article has been cited by 16 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/10/17/5732.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/10/17/5732. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.