British Journal of Cancer (2005) 93, 876 – 883

& 2005 Cancer Research UK All rights reserved 0007 – 0920/05 $30.00
www.bjcancer.com

A Phase I study of the angiogenesis inhibitor SU5416 (semaxanib)

in solid tumours, incorporating dynamic contrast MR
pharmacodynamic end points
Clinical Studies





 A O’Donnell*,1,2, A Padhani3, C Hayes1,2, AJ Kakkar4, M Leach2, JM Trigo1,2, M Scurr1,2, F Raynaud1, S Phillips1,

 W Aherne1, A Hardcastle1, P Workman1, A Hannah5 and I Judson1,2

 1
Institute of Cancer Research, Sutton SM2 5NG, UK; 2Royal Marsden Hospital, Sutton SM2 5PT, UK; 3Paul Strickland Scanner Centre, Mount Vernon

 Hospital, Rickmansworth Road, Northwood, Middlesex, UK; 4Thrombosis Research Institute, Imperial College, London, UK; 5Sugen Inc., South San

 Francisco, CA, USA






 SU5416 (Z-3-[(2,4-dimethylpyrrol-5-yl)methylidenyl]-2-indolinone; semaxanib) is a small molecule inhibitor of the vascular

 endothelial growth factor receptor (VEGFR2). A Phase I dose escalation study was performed. Dynamic contrast-enhanced

 magnetic resonance imaging (DCE-MRI) was used as a pharmacodynamic assessment tool. In all, 27 patients were recruited. SU5416

 was administered twice weekly by fixed rate intravenous infusion. Patients were treated in sequential cohorts of three patients at 48,


 65, 85 110 and 145 mg m
2. A further dose level of 190 mg m
2 after a 2-week lead in period at a lower dose was completed;
 thereafter, the cohort at 145 mg m
2 was expanded. SU5416 showed linear pharmacokinetics to 145 mg m
2 with a large volume of

 distribution and rapid clearance. A significant degree of interpatient variability was seen. SU5416 was well tolerated, by definition a

 maximum-tolerated dose was not defined. No reproducible changes were seen in DCE-MRI end points. Serial assessments of VEGF

 in a cohort of patients treated at 145 mg m
2 did not show a statistically significant treatment-related change. Parallel assessments of

 the impact of SU5416 on coagulation profiles in six patients showed a transient effect within the fibrinolytic pathway. Clinical

 experience showed that patients who had breaks of therapy longer than a week could not have treatment reinitiated at a dose of

 190 mg m
2 without unacceptable toxicity. The 145 mg m
2 dose level is thus the recommended dose for future study.

 British Journal of Cancer (2005) 93, 876 – 883. doi:10.1038/sj.bjc.6602797 www.bjcancer.com


 & 2005 Cancer Research UK


Keywords: antiangiogenic therapy; phase I clinical trial; pharmacokinetics; pharmacodynamics

SU5416 (Z-3-[(2,4-dimethylpyrrol-5-yl)methylidenyl]-2-indoli- confirmed inhibition of tumour growth and a reduction in the

none; semaxanib) has been developed as a potent inhibitor of number of metastases following treatment with SU5416 (Fong et al,
the receptor tyrosine kinase for vascular endothelial growth factor 1999), as well as a decrease in vascular density in treated tumours
receptor – Flk-1/KDR (VEGFR2) (Figure 1). This receptor pathway as measured by intravital video-microscopy (Vajkoczy et al, 1999;
plays a crucial role in the process of angiogenesis. Targeting Vaskoczy et al, 2000).
angiogenesis is highly attractive as a strategy to control Theoretically, an antiangiogenic treatment may prove initially
malignancy. Such a therapy should be less susceptible to the cytostatic rather than cytoreductive. As such, chronic therapy may
development of resistance, since the target is normal endothelial be required, issues of toxicity and compliance become increasingly
cells, and as the majority of normal tissue is not actively important and it would be clinically desirable if one could identify
developing vasculature, it should not be accompanied by the responding patients at the earliest juncture. Thus ideally in
toxicity of conventional cytotoxics. preclinical investigations, but certainly in the Phase I clinical
SU5416 is also a potent competitive inhibitor of KIT (the setting, it is critical to validate noninvasive pharmacodynamic
receptor for stem cell factor) and less potently the PDGF receptor tools that indicate a dose-dependent biological effect.
(indirectly involved in angiogenesis), but is not directly cytotoxic In this Phase I trial, the primary objective was to assess the
by itself (Antonian et al, 2000; Smolich et al, 2001; Yee et al, 2002). safety and tolerability of SU5416 when administered twice weekly
Recent work has also suggested that SU5416 may mediate some of via intravenous infusion. Secondary objectives were to quantify
its effects through inhibition of a second VEGF receptor – Flt-1 effects on tumour vascular permeability using low-molecular
(VEGFR1) (Itokawa et al, 2002). Preclinical xenograft models have weight dynamic contrast-enhanced magnetic resonance imaging
(DCE-MRI), to assess the pharmacokinetics (PK) of SU5416 and to
correlate PK with DCE-MRI. This use of imaging to assess vascular
*Correspondence: Dr A O’Donnell, Wellington Cancer Centre, Well- permeability was designed to investigate SU5416 pharmacody-
ington Hospital, Riddiford Street, Wellington, New Zealand; namics in humans for the first time in a structured dose escalation
E-mail: anne.o’donnell@ccdhb.org.nz manner, in an effort to determine an effective dose for further
Received 23 March 2005; revised 22 August 2005; accepted 22 August development. It is thus unique among parallel Phase I studies with
2005 this compound hitherto reported (Rosen et al, 1999; Stopeck et al,
 Phase I SU5416
A O’Donnell et al
877
(50 mg) as premedication 1 h before treatment, to minimise
hypersensitivity reactions attributable to the excipient Cremo-
phors.
Patients were monitored for at least 1 h after treatment. Samples
for haematology, coagulation studies, biochemistry and urine
N analysis were taken on day 15 and four weekly thereafter. Full
H tumour re-evaluation was performed every 6 weeks using WHO
response criteria. Patients were allowed to continue to receive
therapy with SU5416 indefinitely in the absence of unacceptable
O toxicity or evidence of tumour progression.

N Study drug – SU5416

Clinical Studies
H
Figure 1 SU5416 chemical structure, 3-[(2,4-dimethylpyrrol-5-yl)methy- SU5416 (SUGEN Inc., South San Francisco, CA, USA) was supplied
lideneindolin-2-one. as a yellow-orange liquid formulation in 30 ml vials each contain-
ing 112.5 mg of drug in 25 ml of vehicle for a final concentration of
4.5 mg ml
1. Vehicle constituents in the formulation included:
2002). In addition, effects on coagulation and VEGF levels were polyethylene glycol 400; polyoxyl 35 castor oil (Cremophors); and
explored and any evidence of antitumour activity was reported. benzyl alcohol and dehydrated alcohol. The vials were stored at
room temperature (22 – 281C) and were protected from light. Prior
to administration, SU5416 was diluted 1 : 3 with 0.45% sodium
chloride and injected into non-PVC intravenous infusion bags.
PATIENTS AND METHODS Treatment infusions were administered through non-PVC, non-
Clinical protocol DEHP intravenous tubing and a 0.2 mm in-line filter (not made
from cellulose acetate).
The study protocol was approved by the Research Ethics
Committee of the Royal Marsden NHS Trust. Written informed PK evaluation
consent was obtained from all participants. All patients had
treatment refractory solid tumours and were required to have at Plasma samples for the determination of SU5416 and its
least one tumour deposit suitable for DCE-MRI evaluation. A metabolites were drawn on day 4 and day 25. Samples were taken
suitable lesion had to be X2 cm in size and in an anatomical prior to treatment infusion, 15 min into the infusion, at completion
position not subject to significant respiratory excursion. Eligible of the infusion and then at 5, 10, 20, 30, 45, 60, 120 and 240 min
patients had not received treatment with any other anticancer following the infusion. Blood samples were drawn from the
therapy for 4 weeks prior to treatment with SU5416. Patients were contralateral arm or distant from the venous access device used for
required to have adequate baseline haematological reserves the administration of SU5416. At the designated times, 4.0 ml of
(absolute neutrophil count X1.5  109 l
1, platelet count blood were collected into a lithium heparin tube, gently inverted
X100  109 l
1, haemoglobin X9.0 g dl
1). In addition, all patients and placed on ice to maintain the sample at 2 – 81C. All samples
were required to have adequate systemic organ function (i.e. serum were immediately centrifuged at 6000 r.p.m. at 41C for 10 min, the
transaminases o2.5 times the upper limit of normal; total supernatant divided between two 1.5 ml Nalgene cryovials and
bilirubin o34.2 mmol l
1; and serum creatinine o160 mmol l
1 or snap frozen in liquid nitrogen. Separated plasma samples were
creatinine clearance X50 ml min
1). Exclusion criteria included subsequently stored at
701C until analysis.
patients with a known history of allergy to radiologic contrast Pharmacokinetic analysis was performed using a fully validated
agents or to Cremophors; a Karnovsky performance status of HPLC UV method (Stopeck et al, 2002). Plasma samples, standards
o60%; a history of myocardial infarction, severe or unstable (25 – 2000 ng ml
1) and quality controls spiked with internal
angina within 6 months of the study; diabetes mellitus with clinical standards were extracted with acetonitrile. Following centrifuga-
evidence of severe peripheral vascular disease or ulceration; a tion, drying and reconstitution in 200 ml mobile phase, 20 ml were
medical condition or device implant that would interfere with or analysed by HPLC UV. Chromatography was carried out using a
prevent MRI scans; or a serious concomitant medical condition C18 analytical column, with a gradient mobile phase containing
that could interfere with the participation in the study or acetonitrile and 10 mM ammonium formate at pH 2.6. The flow rate
interpretation of study results. was 0.8 ml min
1 and the run-time 21 min. Detection was via UV
The starting dose of SU5416 for the study was 48 mg m
2. This absorption at a wavelength of 440 nm. The accuracy of the assay,
dose was chosen in the light of available clinical experience in a reflected in relative error of measured concentration to theoretical
concomitant Phase I trial underway in Los Angeles. Sequential (prepared) concentrations, was between 0.9 and 5.6% for all
cohorts of three patients were recruited. Dose escalation steps of quality control samples.
33% were planned. The National Cancer Institute Common Pharmacokinetic parameters were evaluated using WinNonLin
Toxicity Criteria V1 was used to grade toxicity. Dose-limiting Softwares. A comprehensive PK profile of SU5416 and its
toxicity (DLT) was defined as the occurrence of: Grade III or metabolites SU6595 and SU9838 (Cmax, Tmax, T1/2a, T1/2 and Kabs)
greater toxicity excluding nausea/vomiting and haematologic was determined for each patient.
toxicity; or Grade IV haematologic toxicity; or Grade IV nausea
and vomiting refractory to antiemetic therapy. When DLT was Pharmacodynamic evaluations – VEGF
observed in one patient in a cohort, the dose level was expanded to
six patients. The maximum-tolerated dose in this study was Assessment of VEGF levels was incorporated by amendment from
defined as the dose level below which DLT occurred in X2 of the the 145 mg m
2 dose level.
six patients treated in the cohort. Intrapatient dose escalation was Samples for the measurement of VEGF levels were drawn at
permitted. baseline, on day 1 at 2, 4 and 24 h, after the treatment infusion. At
SU5416 was administered twice weekly via intravenous infusion each time point, blood was drawn into a plain tube (serum), a
at a fixed rate of 200 ml h
1. All patients received intravenous sodium citrate tube (platelet-rich plasma) and a CTAD (platelet-
chlorpheniramine (10 mg), dexamethasone (10 mg) and ranitidine stabilised plasma) tube. CTAD tubes (Becton-Dickinson Vacutainer

& 2005 Cancer Research UK British Journal of Cancer (2005) 93(8), 876 – 883
 Phase I SU5416
A O’Donnell et al
878
Systems, Europe) contain four anticoagulants: sodium citrate, RESULTS
theophylline, adenosine and dipyridamole, and are designed to
minimise any platelet activation that may occur as a result of blood Patient characteristics and dose escalation
collection. In all, 27 patients received treatment with SU5416 on this study.
Vascular endothelial growth factor was measured by enzyme- Descriptive details of the patients are shown in Table 1.
linked immunosorbent assay (ELISA) using the commercially Sequential cohorts of three patients were treated at 48, 65, 85,
available Quantikine Kit (R&D Systems, Europe). This analysis was 110, and then 145 mg m
2. In this dose escalation phase, a total of
conducted at The Institute of Cancer Research, Sutton, UK. 24.5 cycles of SU5416 were administered (median per patient, 1.5;
range 0.5 – 4). During the conduct of this trial, investigators in a
parallel Phase I trial of SU5416 on the same schedule at The
Pharmacodynamic evaluation – DCE-MRI
University of California, Los Angeles, USA reported DLT
All patients were asked to have a baseline DCE-MRI examination (projectile vomiting and severe headache) at 190 mg m
2 (Rosen
Clinical Studies

within the 7 days prior to receiving the study compound under et al, 1999). Pharmacokinetic studies in our trial (see below) had
study conditions as described above. Follow-up scans were identified induction in the metabolism of SU5416 occurring within
performed on day 1 and then after 4 weeks of therapy. After the the first weeks of therapy. Therefore, it was decided to conclude
first three patients’ examination, the protocol was amended to day the current study with a final dose expansion cohort of patients
1, 2 weeks, 6 weeks and 6 weekly thereafter if the patient remained receiving 145 mg m
2 twice weekly for four doses, escalating the
on study. All examinations were performed on a 1.5 T MRI system dose to 190 mg m
2 if well tolerated for subsequent therapy. In all,
(Vision, Siemens Medical Systems, Erlangen, Germany) approxi- 12 patients were recruited to this expansion phase of the trial
mately 4 h after the commencement of the treatment infusion. receiving in total 29.5 cycles of therapy (median 2.5; range 0.5 – 5).
The image collection protocol comprised routine orthogonal
T1- and T2-weighted images taken through the previously identified
target lesion, to allow localisation and bi-dimensional tumour Pharmacokinetics
measurement. The dynamic imaging used a five-slice saturation The PK results for the 25 patients evaluable are presented in
recovery turbo fast low-angle shot sequence (SRTF) taken through Table 2. SU5416 was well distributed with a large volume of
the centre of the target lesion. Proton density-weighted images distribution (Vd 39 – 215L). Plasma clearance was rapid (CL 46 –
were acquired, followed by a series of T1-weighted images 215 l h
1), but with large interpatient variability for which there is
employing the same gain and image scaling factors. The sequential no obvious explanation. SU5416 displayed linear kinetics over the
T1-weighted images were acquired every 9 s for 6.3 min (42 time dose range 48 – 145 mg m
2 (Figure 2) at initial assessment and
points). also on day 25. However, the mean AUC at 190 mg m
2 in those
Gadopentetate dimeglumine (Magnevist Schering Health Care patients escalated from 145 mg m
2 to the higher dose after 2
Limited, Burgess Hill, Sussex, UK) was injected intravenously weeks did not differ significantly from that observed at 145 mg m
2
using a mechanical power injector, through a peripherally placed (36117183 vs 37237930 mg l
1 h
1). A similar plateau is observed
cannula. This contrast was administered as a bolus after the third in mean peak concentration. (Table 2) In five of the nine patients
baseline data point was acquired at a dose of 0.1 mmol kg
1 body who were dosed at the same level on both day 4 and day 25, there
weight. The use of the mechanical power injector allowed the was an average increase in drug clearance of 22%.
contrast to be delivered within 10 s. The contrast injection was SU6595 (the carboxyl derivative) was the predominant metabo-
followed by a 20 ml flush of normal saline. lite representing on average 78% of the parent, while SU9838 (the
Using the resultant scanned images, two trained observers (AP, hydroxyl derivative) comprised only 5.2% of the total. Both
CH) blinded to the clinical outcome of the patients evaluated the metabolites behaved as the parent drug with a large volume of
patterns of enhancement within a designated region of interest. distribution and rapid clearance. Neither metabolite shows
Contrast medium kinetic parameters were then derived for each inhibitory activity for any of the receptor tyrosine kinases
pixel using MR Imaging Workbench softwares. Quantitative and inhibited by the parent compound (Mendel et al, 2000). The
semiquantitative statistical analysis was then performed on these increase in clearance observed for the parent compound between
parameters using StatsDirect software (Research Solutions, day 4 and day 25 was not associated with an increase in the plasma
Cambridge, UK). AUC of the main metabolite, SU6595 (data not shown). There was
no correlation between toxicity and any of the PK parameters
evaluated in this study.
Expanded coagulation series
During the dose escalation stage of the protocol, a standard series
of coagulation parameters (APPT, INR and fibrinogen) was
explored at baseline prior to therapy and every 2 weeks thereafter. Table 1 Patient characteristics
In the dose expansion phase of the protocol, in response to the
clinical occurrence of two separate venous thrombotic episodes in Patient characteristics
study patients, an expanded coagulation series was performed.
Age Median 48 years
Samples were taken on cycle 1, day 1 prior to, then at 4 and 24 h Range 18 – 74 years
following the infusion of SU5416 and on cycle 1, day 4 and day 25 Gender 13 males and 14 females
prior to, then 4 h after the SU5416 injection.
At each time point, blood was collected in citrated and EDTA Tumour types
tubes (Becton-Dickinson Vacutainer Systems, Europe). The Soft-tissue sarcoma 10
samples from the citrated tubes were immediately centrifuged (at Ovary/cervix/endometrium 4
41C for 10 min at 3000 r.p.m.) supernatants pooled into inert Melanoma 3
plastic, centrifuged again (at 41C, 10 min at 3000 r.p.m.) and the Renal 2
resultant sample stored in inert plastic tubes at
701C until Head and neck 2
Other (one each) Osteosarcoma, adrenocortical, small cell,
analysis. The sample taken into EDTA was centrifuged (at 41C, for transitional cell, gastro/oesophageal junction,
10 min at 3000 r.p.m.), the supernatant decanted into inert plastic Ca unknown primary
tubes and again stored at
701C until analysis.

British Journal of Cancer (2005) 93(8), 876 – 883 & 2005 Cancer Research UK
 Phase I SU5416
A O’Donnell et al
879
Table 2 Pharmacokinetic results

Patient Dose (mg m
2) Cycle Cmax (lg l
1) AUClast (ha lg l
1) T1/2kz (h) CL (observed) (l h
1) Vz (observed) (l)

1 48 1inj2 1737 463 0.54 181 142

2 48 1inj2 2724 848 0.51 94 69
1inj8 2059 630 0.61 125 109
2inj8 1629 498 0.57 158 131
3 48 1inj2 2253 1143 0.61 83 72
1inj8 1289 816 0.42 115 69
4 65 1inj3 2074 566 0.20 190 55
5 65 1inj2 4371 1498 0.45 64 41
1inj8 4105 1312 0.73 74 78

Clinical Studies
7 65 1inj2 2968 2510 0.82 74 88
1inj8 2914 2068 0.65 89 84
8 85 1inj2 2335 1809 0.39 79 45
1inj8 3672 2375 0.60 60 52
2inj8 2804 1619 0.74 88 94
9 85 1inj2 1624 1151 0.71 133 136
10 85 1inj2 2057 2094 0.71 102 104
1inj8 1354 1250 2.05 151 447
11 110 1inj2 3507 4016 0.62 46 41
12 110 1inj2 2750 2345 0.69 70 70
13 110 1inj2 3237 3842 0.55 83 66
1inj8 3578 3796 0.62 84 75
14 145 1inj2 4851 4946 0.87 55 69
1inj8 3922 4428 0.57 62 51
15 145 1inj2 2869 3471 0.89 92 119
16 145 1inj2 1217 2021 0.71 150 153
1inj8 2080 2349 0.67 169 164
2inj8 1342 1779 0.72 224 233
17 145-190 1inj2 4159 3389 0.68 64 63
1inj8 3370 4467 0.62 63 57
18 145-190 1inj2 2474 1928 0.50 96 70
1inj8 2274 2390 0.50 103 74
3inj2 6517 7173 0.75 44 48
19 145-190 1inj2 2961 3683 0.79 66 75
2inj1 4002 6451 0.77 49 54
20 145-190 1inj2 2774 3397 0.66 89 85
1inj8 2066 2611 0.81 151 176
2inj8 3341 5044 0.72 79 82
21 145-190 1inj2 2894 2621 1.36 100 196
1inj8 2161 2178 2.38 142 488
22 145-190 1inj2 2795 2870 0.67 101 97
23 145-190 1inj2 3202 2465 1.15 72 119
24 145-190 1inj2 1177 1267 0.38 215 118
1inj8 493 816 0.42 352 214
25 145-190 1inj2 4508 4094 0.49 55 39
1inj8 4283 5360 0.58 56 46
27 145-190 1inj2 2625 1578 0.69 106 106

Cmax ¼ maximum concentration; Tmax ¼ time of maximum observed concentration, AUClast ¼ area under the curve from time of dosing to the last measurable concentration;
AUCinf ¼ area under the curve from the time of dosing to infinity; t1/2lz ¼ terminal half-life ln2/lz, where lz is the first-order rate constant associated with the terminal log-linear
portion of the curve; MRTlast ¼ mean residence time from the time of dosing to the last measurable concentration; Vz ¼ volume of distribution based on the terminal phase dose/
lz*AUCinf; CL ¼ total body clearance. aPatient 6 and 26 – not evaluable for PK, samples not taken.

6000 Toxicity
Week 1 Day 4
Week 4 Day 4 With the exception of hypersensitivity, the intravenous infusion
of SU5416 was well tolerated. Despite premedication, Grade II
AUC (h
g l−1)

4000 hypersensitivity reactions affected 11 patients and occurred

independent of the dose of SU5416. All patients were able to
continue to receive therapy with increased corticosteroid cover.
2000 Tachyphylaxis did occur and the steroid dose was able to be
gradually reduced in each patient affected.
Moderate thrombophlebitis was common when peripheral
0 access was used. For those who continued with treatment for
0 50 100 150 200
multiple cycles, central line access was recommended. Other acute
Dose (mg m–2) toxicity included headache, fatigue, nausea and occasional
Figure 2 Pharmacokinetics: AUC vs dose demonstrating linearity and vomiting. Such symptoms were mild, dose related and most
the induction of metabolism between week 1 and week 4. commonly observed in the dose expansion cohort. Of interest,

& 2005 Cancer Research UK British Journal of Cancer (2005) 93(8), 876 – 883
 Phase I SU5416
A O’Donnell et al
880
three patients in this cohort also reported pain or burning Pharmacodynamics
discomfort in tumour-related sites shortly after treatment infu-
sions. Such symptoms were short lived and did not require VEGF VEGF profiles are available for nine patients in the final
intervention. Vital signs were monitored prior to and after cohort. As expected, serum VEGF levels were universally higher
treatment for the duration of the schedule. No predictable or than those obtained from Plasmacitrated or platelet-stabilised
persistent changes were observed. Two patients who received plasma (Plasmactad). A significant correlation between the
therapy for 6 weeks or more developed dysphonia. This did prove VEGF – Plasmactad levels and those in VEGF – Plasmacitrated was
to be reversible over several months. found (r2 ¼ 0.61 Po0.0001); however, there was no comparable
One patient (dose level 145-190 mg m
2, on study 20 weeks) correlation with serum values (Figure 3).
developed recurrent Grade II anaemia that was felt to be related to When the percentage change in VEGF level from baseline was
SU5416. A further patient (dose level 145-190 mg m
2) developed assessed over time, there was a trend suggesting that the levels
a mild (Grade I) elevation in hepatic transaminases, which may dropped initially with subsequent rebound. However, the sample
Clinical Studies

have been due to study therapy, but this patient, with ovarian size was small, the variance wide and the change did not reach
cancer, was known to have a disease plaque adjacent to the porta statistical significance (Figure 4).
hepatis. These were, however, the only haematological and
biochemical drug-related adverse events. DCE-MRI In all, 24 patients were evaluable for changes in
Venous thrombotic events occurred in two patients while they vascular permeability. A total of 80 MRI examinations were
were receiving treatment with SU5416 (thrombosis of the inferior performed in total: 24 patients at baseline; 24 patients on day 1 (3 –
vena cava and a pulmonary embolism). Both patients had serious 4 h post-therapy); 17 patients were scanned mid-treatment (after 2
concomitant risk factors for such medical problems, and these weeks of therapy) and 15 after more than 4 weeks (at a time point
events occurred at differing dose levels, but it is not possible to adjacent to standard tumour evaluation).
exclude an association with SU5416. The criteria for maximum- At baseline, seven patients proved unassessable. Initial tumour
tolerated dose were not met. Of the remaining 26 patients who median transfer constant (Ktrans) was 0.237 min
1 with values
received SU5416 on this study, five were receiving low-dose ranging from 0.091 to 1.325. Median Ve was 32.5% (range 11 –
warfarin for prophylaxis of central venous access lines and one 45%). Pretreatment kinetic parameters did not predict for eventual
patient was receiving therapeutic warfarin for hereditary coagulo- tumour response.
pathy. Nine patients were not evaluable on the first day of treatment.
One patient, with metastatic melanoma (target lesion – brain), had
a statistically significant change in Ktrans, 91% increase. No
Coagulation statistically significant changes in leakage space (Ve) were seen.
No change in INR, APTT or fibrinogen was observed. In all six No dose response was seen in any parameter at this early time
patients evaluated using the expanded panel, a consistent acute point.
increase in plasminogen activator inhibitor (PAI-1) levels was Of the 17 patients evaluated mid-treatment, seven were
observed with recovery at 24 h (median change 37.3 IU; range unassessable. One patient with squamous cell carcinoma and
14.23 – 61 – 27). A parallel reduction in plasmin – anti-plasmin
complexes (median change 12.7 mg l
1; range 0 – 36) was also seen. Correlation between VEGF
There were no consistent changes in the coagulability parameters: sampling techniques
anti-thrombin III, protein C (act), protein S (free), activated 1500
VEGF−CTAD tubes
protein C resistance, dRVVT, endogenous thrombin (generation)
VEGF− serum
potential (extrinsic and intrinsic), prothrombin fragments 1 and 2
VEGF (pg ml−1)

and thrombin – anti-thrombin complexes. No change was seen in 1000 r 2 = 0.61 VEGF CTAD
the endothelial cell haemostatic properties: thrombomodulin or r 2 = 0.008 VEGF serum
Von Willebrand factor (activated and agglutinated).
500

Tumour response
Two patients were not evaluable for response because major 0
intercurrent illness, unrelated to therapy with SU5416 occurring 0 250 500 750 1000
after o2 weeks of therapy, precluded disease assessment VEGF plasma (citrated)(pg ml−1)
(pneumonia, bowel obstruction). In all, 12 patients were con- Figure 3 Correlation between VEGF sampling results taken from
sidered to have stable disease at 6 weeks and went on to receive plasma, serum and platelet-stabilised plasma (CTAD).
further treatment. Of these, three patients remained on therapy for
more than 12 weeks (disease types: haemangioendothelioma,
ovarian and adrenocortical carcinoma). The patient with advanced 100 VEGF levels over time
adrenocortical carcinoma was shown to have a differential
% Change in VEGF from

response with reduction in size of a major metastatic lesion in

50
the retroperitoneum, stable disease in other nodal sites but
progressive disease in the liver. In light of this observation, he
baseline

went on to receive a further 6 weeks of therapy; however, with 0

further evidence of tumour progression in the liver lesions, he was 10 20 30
withdrawn from the study. Progressive disease was the best Time from infusion
response in 14 patients. One of these, a 53-year-old woman with –50 (h)
advanced head and neck carcinoma, was deemed to have
progressive disease having developed necrosis of the nasal area. –100 Mean change at time point
It is possible that this represented a tumour response, but
P = 0.42
palliation was clearly not being achieved. On cessation of therapy, 95% CI
there was an improvement in the degree of tumour-related
erythema and pain. Figure 4 Summary of VEGF changes over time.

British Journal of Cancer (2005) 93(8), 876 – 883 & 2005 Cancer Research UK
 Phase I SU5416
A O’Donnell et al
881
disease in the nasopharynx receiving 110 mg m
2 showed marked deliver 190 mg m
2 on a protracted basis after an introductory
reductions in Ktrans (
51.9%) and Ve (
48.6%). There was no phase of 2 weeks therapy at a lower dose. This dose was previously
evidence of a dose response in any kinetic parameter after 2 weeks determined to cause DLT. In the parallel phase one trial conducted
of therapy. by Rosen et al (1999) using the same schedule as used here, DLTs
The median time to the final scan was 35 days (range 28 – 58 of projectile vomiting, headache and nausea, were reported at
days). Seven of the 15 patients scanned at this time point proved 190 mg m
2. In the dose expansion cohort of our current study
assessable. Two patients had statistically significant increases in (145-190 mg m
2), there were patients who experienced mild
leakage space (36.5 and 34.5%). The first, a patient with soft-tissue headache and emesis. Clinical experience also showed that patients
sarcoma, treated at 48 mg m
2, was found to have progressive who had breaks of therapy longer than a week could not have
disease. The second, again soft-tissue sarcoma, was treated in the treatment reinitiated at a dose of 190 mg m
2 without unacceptable
final expansion cohort, and was shown to have stable disease at toxicity. The 145 mg m
2 dose level was thus considered the
the 6-week assessment. Again there was no evidence of a dose recommended dose for future study.

Clinical Studies
response in the kinetic parameters obtained. Acute hypersensitivity reactions attributable to the diluent
The changes in Ktrans across all dose levels and time points are Cremophors were common. However, in oncology practice such
summarised graphically in Figure 5. reactions are well recognised and can be easily managed. In all
patients in this study, a gradual reduction in prophylactic steroid
dose was possible and did not preclude sustained therapy.
DISCUSSION The overall tolerability of intravenous SU5416, particularly when
central venous access was utilised, was good. Such characteristics
A maximum-tolerated dose of SU5416 when administered as a are important with an antiangiogenic compound such as this,
twice weekly intravenous infusion was not identified in this study. where chronicity of therapy is expected and the toxicity thresholds
Utilising PK observations, we have shown that it is possible to are therefore necessarily set lower.
Homeostasis of coagulation in humans is achieved through a
complex balance of plasma, tissue and endothelial factors. The
100 Mean change in K trans on day 1
process of tumoral angiogenesis mediated through the VEGF
family of growth factors will have effects on vascular permeability
% of Baseline MRI K trans

and endothelial cell proliferation. Thus, we postulated that

75
targeting these pathways with a drug such as SU5416 might result
in physiological consequences within the coagulation pathway.
50 Although the sample size examined was small, a consistent finding
was the observation of an increase in PAI-1 levels in the acute
phase (2 – 4 h after infusion) and a parallel reduction in PAP
25 complexes. These changes reflect a transient effect within the
fibrinolytic pathway. In those patients with additional risk factors,
modulation of PAI-1 and PAP could contribute to an excessive
0 hypercoagulable state and clinical thrombosis. This observation
48 65 85 110 145
may relate to the concomitant administration of dexamethasone
Dose, SU5416 (mg m–2) (Huang et al, 1995; Ha et al, 2002), although a relationship with
SU5416 is also possible. Other than thrombophlebitis directly
75 Mean change in K trans at the end of related to the infusion sites, only two thrombotic episodes, both
week 2 venous, were observed in this study. It is important to note,
% of Baseline MRI K trans

50
however, that almost a quarter of patients received concomitant
warfarin therapy, which even at low dose may have been sufficient
25
to ameliorate this hypercoagulable propensity. An increased
0 incidence of thromboembolic events has been reported in
48 65 85 110 145 190 combination studies with SU5416 (Kuenen et al, 2002a; Cooney
–25 et al, 2005). In affected patients, Kuenen et al (2002b) found an
increase in markers of thrombin generation and endothelial cell
–50 activation, which occurred in a cycle-dependent manner. Discern-
ing the contribution of SU5416 alone to these findings is difficult.
–75 Although increased numbers of thrombotic events are not a feature
of the parallel single-agent Phase I reports (Rosen et al, 1999;
Dose, SU5416(mg m–2) Stopeck et al, 2002), this is perhaps a feature of study size.
75
Mean change in K trans by the end of
Thrombotic events are reported in a number of the single-agent
Phase II studies in the literature to date (Heymach et al, 2004;
% of Baseline MRI K trans

50 week 4
Peterson et al, 2004). Further exploration of this area is warranted
25
and prophylactic anticoagulation is recommended in all combina-
tion studies (Cooney et al, 2005).
0 Pharmacokinetic evaluations confirmed time-dependent induc-
48 65 85 110 145 190 tion of SU5416 metabolism. In this study, the degree of induction
–25 (22%) was less than that seen in the parallel Phase I study (69%)
using the same schedule (Cropp et al, 1999; Rosen et al, 1999). As
–50 the PK sampling for our current study was performed on day 4, in
contrast to day 1 in the Rosen trial, it is likely that the induction
–75 in enzymes of metabolism occurs as early as the first dose. As
Dose, SU5416 (mg m–2) such, the change in clearance mechanisms was probably largely
Figure 5 Dose-related changes in Ktrans using DCE-MRI (displaying mean complete by the time blood sampling was performed during and
and standard deviation). after the second infusion (day 4) in our study. Interestingly, the

& 2005 Cancer Research UK British Journal of Cancer (2005) 93(8), 876 – 883
 Phase I SU5416
A O’Donnell et al
882
degree of induction also appears to relate to frequency of dosing as did not include internal reproducibility or repeatability data for
Stopeck et al (2002) demonstrate a loss of any induction with a the study population. Thus, our ability to determine meaningful
treatment interval of 7 days. The mechanism for the induction in differences in cohorts of this size is curtailed. Thirdly, it is known
SU5416 metabolism is not understood in detail but may relate to that Flk-1-KDR-mediated VEGF activity is unevenly distributed
the upregulation of cytochrome P450 isoenzymes (Antonian et al, within tumour vasculature (Holash et al, 1999; Nakopoulou et al,
2000). The degree of interpatient variability in AUC is tolerable 2002). The ‘regions of interest’ chosen for evaluation in these
given the kinetics are linear and the toxicity profile is favourable. patients may not always have been representative. Finally, the
We believe that the variability is largely driven by interpatient nature of the patient population in this Phase I study may have
differences in handling of SU5416 as the assay has been well influenced the pharmacodynamic results as patients with advanced
validated and results are consistent across trials. As SU5416 in disease may have more mature vasculature that is functionally and
humans is metabolised through liver microsomes to at least six biologically less sensitive to antiangiogenic therapy. As shown by
metabolites (Antonian et al, 2000), again the explanation for the Thomas et al (2001), a homogeneous population with progressive
Clinical Studies

degree of variability is most likely explained by differences in or recently emergent tumours may provide a better opportunity
pharmacogenomics of cytochrome P450 enzymes, affecting the for demonstrating pharmacodynamic relationships than the
metabolism of SU5416 and probably influenced by drug – drug tumours investigated in our study.
interactions. Our results show that SU5416 has a large volume of The evidence of antitumour activity is anecdotal. One patient
distribution and clears rapidly. This is consistent with the with progressive haemangioendothelioma remained on study for 6
assessments in preclinical models (Antonian et al, 2000). Initial months with stable disease and only withdrew because personal
preclinical models also demonstrated that the intracellular half-life circumstances precluded her remaining in the United Kingdom. A
of SU5416 is long and levels capable of target inhibition are second patient, with squamous cell carcinoma of the nasopharynx
maintained for 448 h, with target inhibition resulting in successful developed necrosis of her tumour while receiving treatment. On
inhibition of tumour growth (Fong et al, 1999), thus a twice weekly cessation of therapy, there was clinical improvement in local
infusion schedule was recommended. In vivo experiments, induration and pain. Particularly when viewed in the context of the
however, do not confirm the durability of this inhibition, further marked decrease in vascular permeability observed in this patient
emphasising the importance of developing accurate pharmaco- on MR, this may represent evidence of biological activity with
dynamic markers in the clinical development of such compounds SU5416. In early clinical studies conducted in patients with
(Mendel et al, 2000). tumours characterised by particularly strong signalling through
It is now generally acknowledged that serum samples are the VEGF2 receptor, for example, Kaposi’s sarcoma and haeman-
inappropriate for the measurement of VEGF, as platelets, especially gioblastoma associated with von Hippel – Lindau syndrome,
when activated in clot formation, provide a rich source of this SU5416 has demonstrated encouraging activity (Miles et al, 2000;
growth factor (Maloney et al, 1998; Wynendaele et al, 1999; Aiello et al, 2002; Jennens et al, 2004). In larger Phase II studies
Jelkmann, 2001). The plasma samples obtained in sodium citrate conducted in solid tumours, the response rates have been
and those in the CTAD were reproducible, correlated well and one disappointing (Heymach et al, 2004; Peterson et al, 2004; Stadler
method is not preferred over the other. Although the results did et al, 2004). It seems likely that although the dose level
not reach statistical significance, a trend was suggested with an recommended by the Phase I studies achieves systemic levels that
initial drop in VEGF levels within the first 4 h following infusion, in preclinical modelling produced target and tumour growth
rebounding by 24 h. The timing of this observation suggests it may inhibition (Fong et al, 1999; Mendel et al, 2000), it does so only
reflect the action of dexamethasone (Huang et al, 1995; Ha et al, transiently, and that successful therapeutic angiogenesis inhibition
2002). requires continuously above the threshold for inhibition of a
Dynamic contrast-enhanced MRI offers a unique opportunity critical pathway. Alternatively, the intrinsic redundancy of
for noninvasive functional assessment achievable on most signalling mechanisms associated with the angiogenic process
clinically available MR systems. In a similar study with the VEGFR2 may convey tumour resistance to antiangiogenic compounds
and VEGFR1 inhibitor PTK787/ZK22584 (Thomas et al, 2001; acting mainly on a single target (Kanno et al, 2000; Fox et al, 2001).
Thomas et al, 2002), a correlation was demonstrated between a Further trials using combinations of therapies in this class or
reduction in Ki (equivalent to Ktrans) and both dose and plasma kinase inhibitors with broader target specificities are required.
levels of the study compound. In this study, the tumours studied A twice weekly infusion of SU5416 is safe and able to be
were liver metastases from colorectal cancer and patients at all delivered in a protracted manner. Treatment can achieve
dose levels administered were evaluated using DCE-MRI. potentially therapeutic systemic levels of SU5416. Understanding
New blood vessels in tumours are structurally abnormal; leaky, the PK behaviour of the drug has allowed us to develop an
with fragmented supportive structures (pericytes and basement intrapatient dose escalation schedule (145 mg m
2 twice weekly for
membranes). A successful antiangiogenic therapy would therefore 2 weeks, increasing to 190 mg m
2 if well tolerated), which
result in stabilisation of the vasculature, a reduction in perme- overcomes the lowering of plasma concentration resultant from
ability, levels of interstitial fluid hypertension and thus the size induction of drug metabolism. This trial failed to demonstrate a
of the leakage space (Fox et al, 2001; Jain, 2005). Through PK pharmacodynamic effect using DCE-MRI, but does illustrate some
modelling of the behaviour of a low-molecular weight contrast of the challenges inherent in conducting such studies. Incorpora-
agent, DCE-MRI allows measurement of these parameters: tion of DCE-MRI in further studies with antiangiogenic therapy is
permeability – Ktrans; leakage space – Ve (Galbraith et al, 2002; important for the validation of this tool, but we would suggest that
Choyke et al, 2003). No dose – response changes were seen in Ktrans such studies include an intrapatient baseline assessment of
or Ve at any of the time points assessed in our trial with SU5416. reproducibility and cohort sizes of sufficient number to determine
A failure to demonstrate a consistent change may reflect a number meaningful differences.
of issues. Firstly, and most likely, it may be that SU5416 is
insufficiently potent enough to cause changes in vascular
permeability of sufficient magnitude to be seen with this tool.
Secondly, the sample size was small and as now published the 95% ACKNOWLEDGEMENTS
confidence interval for a meaningful change in Ktrans in a single
patient tumour is a change of
48 to þ 83%, and for Ve 724% This study was supported by SUGEN Inc., South San Francisco,
(Galbraith et al, 2002). It is a limitation of our study that the design CA, USA.

British Journal of Cancer (2005) 93(8), 876 – 883 & 2005 Cancer Research UK
 Phase I SU5416
A O’Donnell et al
883
REFERENCES
Aiello LP, George DJ, Cahill MT, Wong JS, Cavallerano J, Hannah AL, Kuenen BC, Levi M, Meijers JC, Kakkar AK, van Hinsberg VW, Kostense PJ,
Kaelin Jr WG (2002) Rapid and durable recovery of visual function in a Pinedo HM, Hoekman K (2002b) Analysis of coagulation cascade and
patient with von Hippel – Lindau syndrome after systemic therapy with endothelial cell activation during inhibition of vascular endothelial
vascular endothelial growth factor receptor inhibitor SU5416. Ophthal- growth factor/vascular endothelial growth factor receptor pathway in
mology 109(9): 1745 – 1751 cancer patients. Arterioscler Thromb Vasc Biol 22(9): 1500 – 1505
Antonian L, Zhang H, Yang C, Wagner G, Shawver LK, Shet M, Ogilvie B, Maloney JP, Silliman C, Ambruso D, Wang J, Tuder R, Voelkel N (1998)
Madan A, Parkinson A (2000) Biotransformation of the anti-angiogenic In vitro release of vascular endothelial growth factor during platelet
compound SU5416. Drug Metab Dispos 28(12): 1505 – 1512 aggregation. Am J Physiol – Heart Circ Physiol 275(3): H1054 – H1061
Choyke PL, Dwyer AJ, Knopp MV (2003) Functional tumour imaging with Mendel DB, Schreck RE, West DC, Li G, Strawn LM, Tanciongco SS,
dynamic contrast-enhanced magnetic resonance imaging. J Magn Reson Vasile S, Shawver LK, Cherrington JM (2000) The angiogenesis inhi-
Imaging 17: 509 – 520 bitor SU5416 has long lasting effects on vascular endothelial growth

Clinical Studies
Cooney MM, Tserng KY, Makar V, McPeak RJ, Ingalls ST, Dowlati A, factor receptor phosphorylation and function. Clin Cancer Res 6(12):
Overmoyer B, McCrae K, Ksenich P, Lavertu P, Ivy P, Hoppel CL, Remick 4828 – 4858
S (2005) A phase IB clinical and pharmacokinetic study of the angiogenesis Miles S, Arasteh K, Gill P, Jacobs M, Friedman-Kein A, Cropp G, Hannah A
inhibitor SU5416 and paclitaxel in recurrent or metastatic carcinoma of (2000) A multicentre dose-escalating study of SU5416 in AIDS-related
the head and neck. Cancer Chemother Pharmacol 55(3): 295 – 300 Kaposi’s sarcoma. Proc Am Soc Clin Oncol 19: 683
Cropp G, Rosen L, Mulay M, Langecker P, Hannah A (1999) Pharma- Nakopoulou L, Stefanaki K, Panayotopoulou E, Giannopoulou I, Athanas-
cokinetics and pharmacodynamics of SU5416 a phase I, dose escalating siadou P, Gakiopoulou-Givalou H, Louvrou A (2002) Expression of the
trial in patients with advanced solid malignancies. Proc Am Soc Clin vascular endothelial growth factor receptor-2/Flk-1 in breast carcinomas:
Oncol 19: 161a correlation with proliferation. Hum Pathol 33(9): 863 – 870
Fong TA, Shawver LK, Sun L, Tang C, App H, Powell TJ, Kim YH, Schreck Peterson AC, Swiger S, Stadler WM, Medved M, Karczmar G, Gajewski TF
R, Wang X, Risau W, Ullrich A, Hirth KP, McMahon G (1999) SU5416 is (2004) Phase II study of the Flk-1 tyrosine kinase inhibitor SU5416 in
a potent and selective inhibitor of the vascular endothelial growth factor advanced melanoma. Clin Cancer Res 10(12, Part 1): 4048 – 4054
receptor (Flk-1/KDR) that inhibits tyrosine kinase catalysis, tumor Rosen L, Mulay M, Mayers A, Kabbinavar F, Rosen P, Cropp G, Hannah A
vascularization, and growth of multiple tumour types. Cancer Res 59(1): (1999) Phase I dose escalating trial of SU5416, a novel angiogenesis
99 – 106 inhibitor in patients with advanced malignancies. Proc Am Soc Clin
Fox S, Gasparini G, Harris A (2001) Angiogenesis: pathological, prognostic, Oncol 18: 618
and growth factor pathways and their link to trial design and anticancer Smolich BD, Yuen HA, West KA, Giles FJ, Albitar M, Cherrington JM (2001)
drugs. Lancet Oncol 2(5): 278 – 289 The anti-angiogenic protein kinase inhibitors SU5416 and SU06668
Galbraith SM, Lodge M, Taylor J, Rustin GJS, Bentzen S, Stirling J, Padhani inhibit the SCF receptor (c-kit) in a human myeloid leukaemia cell line
A (2002) Reproducibility of dynamic contrast-enhanced MRI in human and in acute myeloid leukaemia blasts. Blood 97(5): 1413 – 1421
muscle and tumours: comparison of quantitative and semi-quantitative Stadler WM, Cao D, Vogelzang NJ, Ryan CW, Hoving K, Wrigth R,
analysis. NMR Biomed 15: 132 – 145 Karrison T, Vokes EE (2004) A randomised phase II trial of the anti-
Ha IS, Um EY, Jung HR, Park HW, Cheong HI, Choi Y (2002) angiogenic agent SU5416 in hormone refractory prostate cancer. Clin
Glucocorticoid diminishes vascular endothelial growth factor and Cancer Res 10(10): 3365 – 3370
exacerbates proteinuria in rats with mesangial proliferative glomerulo- Stopeck A, Sheldon M, Vahedian MG, Cropp G, Gosalia R, Hannah A (2002)
nephritis. Am J Kidney Dis 39(5): 1001 – 1010 Results of a phase I dose escalating study of the antiangiogenic agent,
Heymach JV, Desai J, Manola J, Davis DW, McConkey DJ, Harmon D, Ryan SU5416 in patients with advanced malignancies. Clin Cancer Res 8:
DP, Goss G, Quigley T, Van den Abbeele AD, Silverman SG, Connors S, 2798 – 2805
Folkman J, Fletcher CD, Demetri GD (2004) Phase II study of the Thomas A, Morgan B, Drevs J, Jivan A, Buchert M, Horsfield M, Hennig J,
antiangiogenic agent SU5416 in patients with advanced soft tissue Mross K, Henry A, Ball H, Peng B, Fuxius S, Unger C, O’Byrne K, Laurent
sarcoma. Clin Cancer Res 10(17): 5732 – 5740 D, Dugan M, Steward W (2001) Pharmacodynamic results using dynamic
Holash J, Wiegand SJ, Yancopoulos GD (1999) New model of tumor contrast enhanced magnetic resonance imaging of 2 phase 1 studies of
angiogenesis: dynamic balance between vessel regression and growth the VEGF inhibitor PTK787/ZK 222584 in patients with liver metastases
mediated by angiopoietins and VEGF. Oncogene 18(38): 5356 – 5362 from colo-rectal cancer. Proc Am Soc Clin Oncol 20: 279
Huang LQ, Whitworth JA, Chesterman CN (1995) Effects of cyclosporin A Thomas A, Morgan B, Decatris A, Henry A, Pfister C, Souppart C, Gano C,
and dexamethasone on haemostatic and vasoactive functions of vascular Moss D, Laurent D, Dugan M, Steward W (2002) A phase I study of
endothelial cells. Blood Coagul Fibrin 6(5): 438 – 445 the oral vascular endothelial growth factor (VEGF) receptor tyrosine
Itokawa T, Nokihara H, Nishioka Y, Sone S, Iwamoto Y, Yamada Y, kinase inhibitor PTK787/ZK222584 on a twice-daily schedule in patients
Cherrington J, McMahon G, Shibuya M, Kuwano M, Ono M (2002) with advanced cancer. Proceedings of the NCI/AACR/EORTC Confe-
Antiangiogenic effect by SU5416 is partly attributable to inhibition of rence November 2002, Amsterdam: Molecular Targets and Cancer
Flt-1 receptor signalling. Mol Cancer Therap 1(5): 295 – 302 Therapeutics, Abstract 244
Jain RK (2005) Normalization of tumor vasculature: an emerging concept Vajkoczy P, Menger MD, Vollmar B, Schilling L, Schmiedek P, Hirth KP,
in antiangiogenic therapy. Science 307: 58 – 62 Ullrich A, Fong TA (1999) Inhibition of tumor growth, angiogenesis, and
Jelkmann W (2001) Pitfalls in the measurement of circulating vascular microcirculation by the novel Flk-1 inhibitor SU5416 as assessed by
endothelial growth factor. Clin Chem 47: 617 – 623 intra-vital multi-fluorescence video microscopy. Neoplasia 1(1): 31 – 41
Jennens RR, Rosenthal MA, Lindeman GJ, Michael M (2004) Complete Vaskoczy P, Thruner A, Hirth KP, Schilling L, Schmiedek P, Ullrich A,
radiological and metabolic response of metastatic renal cell carcinoma to Menger MD (2000) Measuring VEGF-Flk-1 activity and consequences of
SU5416 (semaxanib) in a patient with probable von Hippel – Lindau VEGF-Flk-1 targeting in vivo intravital microscopy: clinical applications.
syndrome. Urol Oncol 22(3): 193 – 196 Oncologist 5(Suppl 1): 16 – 19
Kanno S, Oda N, Abe M, Terai Y, Ito M, Shitara K, Tabayashi K, Shibuya M, Wynendaele W, Derua R, Hoylearts MF, Pawinski A, Waelkens E, de Bruijn
Sata Y (2000) Roles of two VEGF receptors, Flt-1 and KDR, in the signal EA, Paridaens R, Merlevede W, van Oosterom A (1999) Vascular
transduction of VEGF effects in human vascular endothelial cells. endothelial growth factor measured in platelet poor plasma allows
Oncogene 19(17): 2138 – 2146 optimal separation between cancer patients and volunteers: a key to
Kuenen BC, Rosen L, Smit E, Parson M, Levi M, Ruijter R, Huisman H, study an angiogenic marker in vivo? Ann Oncol 10: 965 – 971
Kedde MA, Noordhuis P, van der Vijgh W, Peters G, Cropp G, Scigalla P, Yee KW, O’Farrell AM, Smolich BD, Cherrington JM, McMahon G,
Hoekman K, Pinedo HM, Giaccone G (2002a) Dose finding and Wait CL, McGreevey LS, Griffith DJ, Heinrich MC (2002) SU5416 and
pharmacokinetic study of cisplatin, gemcitabine and SU5416 in patients SY5614 inhibit kinase activity of wild-type and mutant FLT-3 receptor
with solid tumors. J Clin Oncol 20(6): 1657 – 1667 tyrosine kinase. Blood 100(8): 2941 – 2949

& 2005 Cancer Research UK British Journal of Cancer (2005) 93(8), 876 – 883