original article Annals of Oncology 21: 2220–2226, 2010

doi:10.1093/annonc/mdq221
Published online 28 April 2010

Treatment-emergent hypertension and outcomes in

patients with advanced non-small-cell lung cancer
receiving chemotherapy with or without the vascular
endothelial growth factor receptor inhibitor cediranib:
NCIC Clinical Trials Group Study BR24 
R. Goodwin, K. Ding, L. Seymour, A. LeMaı̂tre, A. Arnold, F. A. Shepherd, M. Dediu,
T. Ciuleanu, D. Fenton, M. Zukin, D. Walde, F. Laberge, M. Vincent, P. M. Ellis & S. A. Laurie*
from the NCIC Clinical Trials Group, Kingston, Ontario, Canada
NCIC Clinical Trials Group, Queen’s University, Kingston, Ontario, Canada

Received 17 December 2009; revised 5 March 2010; accepted 8 March 2010

Background: Hypertension (HTN), a recognized adverse effect of angiogenesis inhibitors, may be a potential
biomarker of activity of these agents. We conducted a retrospective analysis to examine the incidence and predictors
of the development of on-treatment HTN with the vascular endothelial growth factor receptor tyrosine kinase inhibitor
cediranib, and the relationship of this adverse event with treatment outcomes.
Patients and methods: BR24 was a double-blind placebo-controlled phase II trial of carboplatin/paclitaxel
original
article

chemotherapy with either daily oral cediranib or placebo in patients (n = 296) with advanced non-small-cell lung cancer
(NSCLC). Exploratory analyses characterized relationships between HTN, baseline variables, and efficacy outcomes.
Results: New onset or worsening of preexisting HTN (treatment-emergent HTN) was more frequent in patients
receiving cediranib (68 versus 45%, P < 0.0001). Factors associated with HTN in all randomized patients were good
performance status and treatment with cediranib. In both arms, treatment-emergent HTN was associated with
improved efficacy outcomes, but there was no evidence of a differential treatment effect, with nonsignificant interaction
P values.
Conclusions: In advanced NSCLC, HTN is frequent in patients receiving chemotherapy, with or without cediranib.
The development of HTN was favorably prognostic in these patients, but not predictive of a differential outcome with
cediranib.
Key words: angiogenesis inhibitor, biomarker, predictive marker

introduction increased survival in selected patients with advanced non-

Tumor growth and metastasis are dependent on new blood
vessel formation, which is driven by vascular endothelial
growth factor (VEGF) [1]. Inhibition of angiogenesis is an
important new treatment strategy. Angiogenesis inhibitors (AI)
targeting the tyrosine kinases (TK) of the VEGF receptors
(sunitinib and sorafenib) are approved for hepatocellular [2]
and renal cell carcinomas [3, 4]. In an Eastern Cooperative
Oncology Group (ECOG) study, bevacizumab, a monoclonal
antibody against VEGF plus carboplatin/paclitaxel (C+P)
*Correspondence to: Dr S. A. Laurie, Division of Medical Oncology, Ottawa Hospital
Cancer Centre, University of Ottawa, 501 Smyth Road, Ottawa Ontario, Canada
K1H 8L6. Tel: +1 613-737-7700 ext 70175; Fax: +1 613-247-3511;
E-mail: slaurie@ottawahospital.on.ca
Presented in part at the Annual Meeting of the American Society of Clinical Oncology,
Orlando, Florida, May 2009.
small-cell lung cancer (NSCLC) [5], although a survival
advantage was not observed when combined with cisplatin/
gemcitabine [6].
Biomarkers of clinical benefit that help to improve selection
of therapy for patients include pretreatment predictive factors,
such as mutations in the epidermal growth factor receptor
[EGFR] for EGFR inhibitors. It may also be possible to identify
early, treatment-emergent surrogates of efficacy, allowing
patients who are unlikely to respond to discontinue therapy.
The development of rash in patients treated with the EGFR
inhibitor erlotinib was associated with a survival benefit in
patients with NSCLC and pancreatic carcinoma [7, 8], and in
patients with NSCLC receiving cetuximab with chemotherapy
[9]. There are no validated predictive biomarkers for selecting
patients for AI therapy [10] nor are there any arising during
therapy that serve as surrogates for outcome.

ª The Author 2010. Published by Oxford University Press on behalf of the European Society for Medical Oncology.
All rights reserved. For permissions, please email: journals.permissions@oxfordjournals.org
Annals of Oncology
Hypertension (HTN) is a known adverse event (AE) of AIs,
although its mechanism is unclear. VEGF induces vasodilation
by stimulating nitric oxide release from endothelial cells [11].
In patients treated with sorafenib, measures of vascular stiffness
increased [12]; VEGF inhibition may lead to vasoconstriction
and increased peripheral vascular resistance. Reduced
microvessel density (rarefaction) has been proposed as
a potential cause [13, 14]. Given its high incidence, HTN may
be a useful pharmacodynamic surrogate of outcome.
NCIC Clinical Trials Group (NCIC CTG) study BR.24 was
a randomized double-blind phase II trial of daily oral cediranib
or placebo in combination with C+P in advanced NSCLC.
Cediranib (AZD2171, Recentin
; AstraZeneca, Macclesfield,
UK) is a potent inhibitor of the TK activity of all VEGF
receptors. In the primary phase II efficacy analysis, cediranib 30
mg daily with chemotherapy increased response rate (RR; 38%
versus 16%, P < 0.0001) and progression-free survival (PFS;
hazard ratio [HR] 0.77, 95% confidence interval [CI] 0.56–1.08,
P = 0.13) over C+P alone [15]; in a final analysis of all 296
patients (45 and 30 mg), a similar HR for overall survival (OS)
was seen (HR 0.78, 95% CI 0.57–1.06, P = 0.11).
This retrospective exploratory analysis examined the
incidence and predictors of HTN in BR.24, its association with
other AE and effects on drug delivery, and its potential
surrogate effect for efficacy.
patients and methods
study population
Patients with previously untreated stage IIIB/IV NSCLC received
carboplatin (area under the curve = 6) and paclitaxel (200 mg/m2) every
3 weeks for six to eight cycles, with daily oral cediranib/placebo, which
was continued as monotherapy after chemotherapy completion in the
absence of intolerable toxicity or disease progression. The initial cediranib/
placebo dose was 45 mg/day (n = 45) with an early amendment to
30 mg/day (n = 251) due to toxicity concerns. The main efficacy results
have been reported previously [15].
Patients were assessed every cycle for toxicity, graded using Common
Toxicity Criteria Adverse Event (CTCAE), version 3.0 (National Cancer
Institute, Bethesda, MD), and every 6 weeks for response by RECIST [16].
blood pressure measurements
Patients with a history of HTN were eligible provided that at enrolment
blood pressure (BP) was <150 mmHg systolic and <100 mmHg diastolic,
and there was no history of poorly controlled or labile HTN, or poor
compliance with antihypertensive medications. BP was recorded weekly by
a health professional for cycles 1–3, then every 3 weeks until off protocol.
HTN toxicity and management
For this analysis, HTN was defined as either new-onset HTN or worsening
grade (CTCAE v3.0) from baseline in a patient with a past history of HTN
using AE data and actual BP measurements. For preexisting HTN, any
increase in drug dosage or initiation of a new antihypertensive agent was
called grade 3.
An algorithm developed by NCIC CTG during phase I studies of this
agent [17] suggested antihypertensive agents to initiate or add, given the
purported mechanism of HTN and preclinical data regarding cediranib-
associated HTN [18], while minimizing both potential drug interactions
with cediranib, and dose reductions and interruptions of cediranib/placebo.
Guidelines on holding, dose reducing, or discontinuing cediranib/placebo
Volume 21 | No. 11 | November 2010
original article
were given. Briefly, cediranib/placebo was discontinued for grade 4 HTN or
for uncontrollable moderate (150–179/100–109) to severe but
asymptomatic (>180/>110) HTN. Cediranib/placebo was held only for
severe asymptomatic HTN or if moderate HTN was not controlled after
the addition of two drugs; in either of these situations, if restarted, the dose
was reduced.
statistics
Exploratory analyses characterized the relationships between HTN and
baseline variables (age, gender, ECOG performance status [PS], stage,
smoking history, histology, lactate dehydrogenase [LDH], albumin, past
history of HTN, body surface area [BSA]) and efficacy outcomes. Chi-
square test or Fisher’s exact test was used to assess association between
categorical variables, and logistic regression model was used to identify
factors predicting for the development of HTN.
A Cox regression model with HTN as a time-dependent covariate
examined whether HTN could predict the outcomes of OS and PFS in both
univariate and multivariate analyses. The OS was defined as from the
time of randomization to the date of death or was censored at the last
known date alive.
Landmark analyses were also performed, with the time-points for the
development of HTN at either the end of cycles 1 or 2, to see if early-onset
HTN, before the first response assessment, predicted for outcome, and
included all patients who were alive at the time-point. These time-points
represented roughly the median time to onset of HTN (cycle 1), and the
time of the first response assessment (cycle 2), which was also a point at
which the majority of HTN events had occurred in each arm. Kaplan–Meier
curves were used to summarize distributions of the time-to-event
outcome and compared with the log-rank test.
results
patient population
Of 296 patients accrued, 294 who received at least one dose
of cediranib/placebo are included in this analysis (Table 1).
The study population was generally balanced although more
cediranib patients were age >60 (52% versus 40%, P = 0.05) and
had adenocarcinoma (57% versus 45%, P = 0.08). The arms
were balanced for factors that could potentially influence the
development of HTN, such as a past history of HTN, body
mass index, and creatinine clearance.
predictors of HTN
In univariate analyses, ECOG PS of 0 predicted HTN in both
the cediranib (P = 0.02) and placebo groups (P = 0.008;
Table 2). For cediranib, female sex (P = 0.004), normal LDH
(P = 0.015), and no prior peripheral vascular disease (P =
0.015) predicted HTN; in the placebo arm, nonsquamous
histology (P = 0.006), age > 60 (P = 0.018), and a history of
diabetes (P = 0.046) also predicted HTN. Among all 294
patients, factors that independently predicted HTN were
treatment with cediranib, good PS, and a normal LDH; neither
a past history of HTN nor smoking status (ever versus never
smoking or current smoking) predicted for HTN.
incidence of HTN
Cediranib patients were more likely to develop any grade HTN
(68% versus 45%, P < 0.0001), as well as grade 3/4 HTN
(16% versus 3%, P < 0.001; Tables 3 and 4). The median time
doi:10.1093/annonc/mdq221 | 2221
original article Annals of Oncology

Table 1. Baseline characteristics of study population by arm

Placebo (all patients, Cediranib (all patients, Cediranib 30 mg/day Cediranib 45 mg/day
n = 146), n (%) n = 148), n (%) (n = 126), n (%) (n = 22), n (%)
Age, median (range) 58 (39–81) 61 (35–82) 60 (35–77) 64 (42–82)
Gender, female 61 (41) 62 (42) 53 (42) 9 (41)
Stage, n (%)
IIIB 23 (16) 19 (13) 16 (13) 3 (14)
IV 123 (84) 129 (87) 110 (87) 19 (86)
ECOG, n (%)
0 39 (27) 33 (22) 26 (21) 7 (32)
1 or 2* 107 (73) 115 (78) 100 (79) 15 (68)
Histology, n (%)
Adeno 65 (45) 85 (57) 74 (59) 11 (50)
Squamous 37 (25) 27 (18) 24 (19) 3 (14)
Other 44 (30) 36 (24) 28 (22) 8 (36)
History of HTN, n (%) 48 (33) 39 (26) 32 (25) 7 (32)
BSA, median 1.80 1.82 1.80 1.86
BMI, median 24.95 24.52 24.34 24.86
Creatinine clearance, median 90.93 89.94 91.19 84.80
*
ECOG 2 patients were only included in the cediranib/placebo 45 mg/day cohort (n = 45).
ECOG, Eastern Cooperative Oncology Group; HTN, hypertension; BSA, body surface area; BMI, body mass index.

Table 2. Multivariate analysis of HTN adverse event predictors in all patients on cediranib, headache was the only symptom
patients correlating with HTN (P = 0.007). For patients on placebo,
no correlations were found.
Odds ratio 95% CI P value
ECOG > 0 0.3 0.1–0.5 0.0001 HTN as a surrogate of treatment effect
Increased LDH 0.5 0.3–1.0 0.03 When the relationship between efficacy and HTN as a time-
Receiving cediranib 2.9 1.7–4.9 0.0001 dependent covariate was examined, it revealed that for
cediranib patients, developing HTN reduced the risk of death
Other factors in the model, which were not predictive were gender, stage,
by 38% (HR new/worsening HTN versus without: 0.62, 95%
histology, smoking history, albumin (normal versus low), body surface area
CI 0.38–1.03, P = 0.06), while for placebo patients, it reduced
(<2 versus ‡2), histories of HTN/cerebrovascular disease/renal disease/
the risk of death by 51% (HR 0.49, 95% CI 0.30–0.80, P =
diabetes/coronary artery disease/hypercholesterolemia. CI, confidence
interval; ECOG, Eastern Cooperative Oncology Group; LDH, lactate
0.0045). Multivariable Cox regression analysis of all 294
dehydrogenase. patients showed that HTN independently predicted OS (HR
0.7, 95% CI 0.5–1.0, P = 0.06). In cediranib patients, HTN
to onset was 13.5 days in the cediranib arm (range 1–168 days) was not significantly associated with RR (42 versus 35%, P =
and 18.5 days in the placebo arm (range 1–200 days); the 0.48) or with significantly improved PFS (HR 0.78, P = 0.28).
median time to onset of grade 3/4 HTN was 16 days in the Similar results for RR (27 versus 13 %, P = 0.03) and PFS
cediranib-treated patients (range 1–65 days). While there were (HR 0.87, P = 0.49) were observed for those patients on
numerically more patients with grade 1/2 HTN at 45 mg of placebo. The interaction-by-treatment arm statistics were
cediranib, there were no episodes of severe HTN in this cohort nonsignificant for RR (P = 0.23), OS (P = 0.73), and PFS
and there was no clear evidence of a dose response for HTN, (P = 0.66).
likely due to small sample size. Landmark analyses (Table 5), performed at the two time
points, gave similar results and were consistent with the results
of the time-dependent covariate Cox regression model,
impact on drug delivery
demonstrating increased OS among patients with HTN in both
Effects on drug delivery were minimal. HTN was the cause arms, without evidence of a differential treatment effect.
of reduction or discontinuation of cediranib/placebo in only Figure 1 displays the Kaplan–Meier curves for the landmark
11 patients (9 cediranib, 2 placebo); only one patient analyses.
interrupted chemotherapy.

sequelae of HTN discussion

Correlations between HTN and other relevant AEs such as HTN, a mechanistic effect of AI, is seen with both bevacizumab
headache, intracranial bleeding, abnormal fundoscopy, [19] and VEGFR TK inhibitors [20, 21]. HTN occurs and peaks
proteinuria, and worsening creatinine were examined. For early (within the first 3 weeks of therapy) with the TK

2222 | Goodwin et al. Volume 21 | No. 11 | November 2010

Annals of Oncology original article
Table 3. Worst grade of HTN reported on study by treatment arm

Grade Placebo (all patients, Cediranib (all patients, Cediranib 30 mg/day Cediranib 45 mg/day
HTN n = 146), n (%) n = 148), n (%) (n = 126), n (%) (n = 22), n (%)
0 80 (55) 48 (32) 44 (35) 4 (18)
1/2 62 (42) 77 (52) 59 (47) 18 (82)
3 4 (3) 22 (15) 22 (17) 0
4 0 1 (1) 1 (1) 0

Table 4. New onset HTN or worsening HTN grade by treatment arm

Placebo (all patients, Cediranib (all patients, Cediranib 30 mg/day Cediranib 45 mg/day
n = 146), n (%) n = 148), n (%) (n = 126), n (%) (n = 22), n (%)
New onset HTN 40 (27) 72 (49), P = <0.0001* 60 (48) 12 (55)
Worsening HTN grade 26 (18) 28 (19) 22 (17) 6 (27)
New onset HTN or 66 (45) 100 (68), P < 0.0001* 82 (65) 18 (82)
worsening HTN grade
*
P value is compared with placebo cohort.
HTN, hypertension.

Table 5. Landmark analyses

Incidence, Response Median PFS Median survival

HTN (%) rates (months) (months)
End cycle 1
Cediranib (n = 144) 40 +HTN 40% Interaction 5.7 Interaction 10.7 Interaction
2HTN 41%, P = 1.0 P = 0.31 5.7, HR 0.69, P = 0.37 8.5 HR 0.62, P = 0.94
P = 0.09 P = 0.07
Placebo (n = 146) 27 +HTN 26% 5.3 15.9
2HTN 17%, P = 0.24 5.0, HR 0.94, 8.0, HR 0.62,
P = 0.76 P = 0.09
End cycle 2
Cediranib (n = 138) 58 +HTN 40% Interaction 6.6 Interaction 11.4 Interaction
2HTN 47%, P = 0.49 P = 0.27 5.7, HR 0.63, P = 0.35 7.7, HR 0.51, P = 0.99
P = 0.03 P = 0.007
Placebo (n = 143) 33 +HTN 23% 5.5 15.9
2HTN 18%, P = 0.5 4.9, HR 0.85, 7.5, HR 0.52,
P = 0.44 P = 0.02

HTN, hypertension; HR, hazard ratio; PFS, progression-free survival.

inhibitors [12, 22]; the median time to onset and to maximal
BP increase with bevacizumab is later [23, 24]. In the phase I
trial of single-agent cediranib, HTN was a dose-limiting toxicity
in nine patients [25]. In phase I studies of cediranib in
combination with cisplatin/gemcitabine [26] and C+P [17],
grade 3/4 HTN was observed in 7 of 15 and 7 of 20 patients,
respectively, and led to the development of the treatment
algorithm used in this protocol.
In BR.24, new or worsening HTN was observed in
approximately two-thirds of patients receiving cediranib, with
a median time of onset of 2 weeks. Predictors of cediranib-
induced HTN included female gender and good PS, which
were also predictors of bevacizumab-induced HTN in the
ECOG trial [27]. Although both good PS and female gender
are themselves favorably prognostic in NSCLC, it is unlikely
that the greater incidence is related to greater cediranib
exposure since the majority of patients developed HTN before
their first response assessment; furthermore, the multivariate
analysis adjusted for HTN as a time-varying covariate.
Cediranib dosing is not body weight based, but HTN was not
associated with BSA, and bevacizumab is dosed by weight.
Cediranib pharmacokinetics did not differ by sex. Thus,
greater cediranib exposure due to fixed dosing does not likely
account for the increased risk in women, and the reason why
women may be inherently more susceptible to HTN from AI is
unknown.
Treatment-emergent HTN was also observed in 45% of those
receiving placebo. This comparatively high rate of documented
HTN is likely due to the frequent protocol-mandated
monitoring, the fact that data were abstracted centrally based

Volume 21 | No. 11 | November 2010 doi:10.1093/annonc/mdq221 | 2223

original article Annals of Oncology

NCIC CTG TRIAL BR.24 NCIC CTG TRIAL BR.24

Overall survival in AZD2171 arm Overall survival in placebo arm
100 100
HTN No HTN HTN No HTN

80 80
Percentage

Percentage
60 60

40 40

20 20

0
0
0 4 8 12 16
Time(Months) 0 4 8 12 16
# at Risk Time(Months)
No HTN 87 58 30 11 2 # at Risk
57 34 19 9 5 No HTN 107 72 32 12 4
HTN
HTN 39 30 15 9 3
Summary Statistics:
Test for equality of groups: Log Rank: p=0.0704 Summary Statistics:
Test for equality of groups: Log Rank: p=0.0863
Median (95%C.I.): No HTN: 8.5 (7.6,10.3), HTN: 10.7 (7.9,21.8)
Median (95%C.I.): No HTN: 8.0 (6.6,10.5), HTN: 15.9 (6.3,24.6)
Hazard Ratio (95%C.I.): HTN/No HTN: 0.62 (0.37,1.05)
Hazard Ratio (95%C.I.): HTN/No HTN: 0.62 (0.35,1.08)

NCIC CTG TRIAL BR.24

NCIC CTG TRIAL BR.24 Overall survival in placebo arm
Overall survival in AZD2171 arm
100 HTN No HTN
100
HTN No HTN

80 80
Percentage

60
Percentage

60

40
40

20
20

0
0 4 8 12 16 0
# at Risk Time(Months)
58 40 17 7 2
0 4 8 12 16
No HTN Time(Months)
HTN 80 52 32 13 5 # at Risk
No HTN 96 65 28 11 4
Summary Statistics: HTN 47 37 19 10 3
Test for equality of groups: Log Rank: p=0.0074
Median (95%C.I.): No HTN: 7.7 (6.9,10.1), HTN: 11.4 (8.8,21.8) Summary Statistics:
Hazard Ratio (95%C.I.): HTN/No HTN: 0.51 (0.31,0.84) Test for equality of groups: Log Rank: p=0.0185
Median (95%C.I.): No HTN: 7.5 (6.3,10.5), HTN: 15.9 (8.0,24.6)
Hazard Ratio (95%C.I.): HTN/No HTN: 0.52 (0.30,0.91)

Figure 1. Landmark analyses: Overall survival for patients with and without HTN, by treatment arm. Top panels: Landmark analyses end of cycle 1. Bottom
panels: Landmark analyses end of cycle 2.

on individual measurements and to underreporting in other
trials. HTN arising during therapy with standard cytotoxics has
not been well described. In a previous NCIC CTG trial (BR.18),
the same C+P regimen led to a rate of HTN reported as an
AE of only 1% [28]. More recently, others have shown that
HTN occurs in patients receiving standard chemotherapy
alone, albeit at a lower rate and of lesser severity than when
combined with an AI [29–31]. In BR.24, the incidence and
severity were both significantly less on the placebo arm, with
only 3% developing grade 3 HTN comparable with the rate of
grade 3/4 HTN documented in the C+P alone arm of other
trials evaluating AI in advanced NSCLC [5, 32]. The possibility
that a patient may be receiving an AI potentially accounts for
the increased reporting of HTN as an AE in those receiving
chemotherapy alone in a blinded trial. HTN did not lead to
significant reductions in treatment delivery, suggesting that
oncologists are comfortable managing this class-specific
toxicity, and confirming the usefulness of the algorithm.
Prompt institution of therapy likely prevents more severe
HTN, and may decrease the risk of other AE described with
AI that may also be partly mediated by HTN. This includes
congestive heart failure, myocardial ischemia, reversible
posterior leukoencephalopathy syndrome, and cerebrovascular
hemorrhage and ischemia, none of which were observed in
BR.24. Cediranib-induced HTN was not associated with other
AE except headache, suggesting that this symptom should
prompt an assessment of BP in a patient receiving AI.
An association between treatment-emergent HTN and
improved outcomes with bevacizumab and VEGFR TK
inhibitors has been observed in a variety of tumor types,
including colorectal [33], pancreatic [34], and renal cell
carcinomas [35, 36]. These reports of patients treated with
an AI (with or without chemotherapy) have compared those
who develop HTN with identically treated patients who do not.
Similar findings have been observed with bevacizumab in
NSCLC. An analysis of patients enrolled to the bevacizumab

2224 | Goodwin et al. Volume 21 | No. 11 | November 2010

Annals of Oncology
arm of the ECOG study suggested increased OS and PFS in
those with HTN compared with those without HTN [27].
In a review of over 2000 patients treated with bevacizumab in
conjunction with a variety of chemotherapies for advanced
NSCLC, median survival was 18.8 months in those who
developed HTN and 12.9 months in those who did not [37].
Analyses of single-arm studies, or of only the AI arm of
a randomized study, do not distinguish whether HTN is
predictive of benefit from AI, or if HTN is simply prognostic.
To our knowledge, our analysis is the first to evaluate the
potential surrogate biomarker effect of HTN in both arms of
a randomized study. While limited because it was a randomized
phase II with a modest sample size and for which OS was not
the primary end point, the trial was blinded, patients had
frequent BP monitoring, and careful recording of all
antihypertensive/concomitant medications, and thus the data
are robust. We have shown that those patients who developed
HTN had better efficacy outcomes, but this association was,
surprisingly, in both arms. The magnitude of the effect, as
measured by HRs, was similar in each arm, and the interaction
P values were not significant. Although the sample size was
small, there was not even a trend in favor of a treatment effect,
and the results of all analyses were similar. As there was no
OS difference between the two study arms in this trial, one
cannot conclusively say there is no differential surrogate
effect of HTN on survival; however, the lack of observed
differential effect in RR or PFS makes it unlikely that HTN is
a surrogate for efficacy.
We performed two different types of analyses to evaluate the
relationship between HTN and outcome to try to obviate the
potential bias introduced by the fact that HTN will occur at
different times, and that patients who continue on therapy
longer because they are benefiting from treatment have
a greater chance of developing HTN. The time-dependent
covariate analysis includes and classifies all patients
appropriately, but it is possible that if early dropouts or deaths
occur before the onset of HTN, this would make the no-HTN
arms look worse, and give rise to an artificial association
between HTN and favorable outcome. The landmark analyses
could also be biased by these early deaths, and by definition
misclassifies those patients who develop HTN after the time-
point used. Nevertheless, within these limitations, our data
do not support the use of the development of HTN as
a surrogate marker for benefit from cediranib but instead
suggest that HTN is favorably prognostic in advanced NSCLC
treated with C+P.
This finding of improved outcome in patients with advanced
NSCLC who develop HTN while on treatment is intriguing.
Why this would be is unclear and requires further investigation.
Paclitaxel has been purported to have antiangiogenic properties
[38], with inhibition of endothelial cell proliferation and
motility. This could explain some of the HTN observed with
C+P, and may obscure the ability to discern a difference
between the control and experimental arms of this trial.
Analyses of other clinical trials that utilize chemotherapies for
which no antiangiogenic activity has been described would be
informative. Alternatively, the reduction in tumor burden in
responding patients may lead to reduced systemic circulating
levels of VEGF which could result in the development of HTN.
Volume 21 | No. 11 | November 2010
original article
Correlating HTN and outcomes with on-treatment changes
in serum VEGF and other markers of angiogenesis could help
to elucidate if this hypothesis was plausible. The VEGF gene is
highly polymorphic, and variants that appear to be associated
with a decreased risk for HTN were detected in a trial of
paclitaxel and bevacizumab in advanced breast cancer [39];
these were not the same alleles that were favorably prognostic,
and in that trial, HTN was associated with OS on the
bevacizumab arm. An imbalance in these polymorphisms
between treatment arms may account for differences in rates of
HTN and the unexpected findings between HTN and outcome
in BR.24; an analysis of patient samples is ongoing. Finally,
HTN may be a marker for an as yet unidentified surrogate for
chemotherapy efficacy.
These data suggest that treatment-emergent HTN is common
not only in patients with advanced NSCLC receiving systemic
therapy with an AI, but also in those receiving a standard
chemotherapy, and may be underreported. Since HTN occurs
early with TK inhibitors, frequent monitoring for the first
cycles of therapy, and also if patients develop headache, seems
prudent, to permit continuation of AI therapy, and to prevent
more serious HTN-related consequences. The development
of HTN on therapy appears to be favourably prognostic, but
did not predict benefit from cediranib in this clinical trial.
funding
The NCIC CTG is supported by the Canadian Cancer Society.
AstraZeneca provided partial funding and cediranib/placebo
for BR.24.
disclosures
FAS, PME and MD have recived honoraria from AstraZeneca.
LS has research funding and stock ownership from
AstraZeneca. RG, KD, AL, AA, TC, DF, MZ, DW, FL, MV, and
SAL have no conflicts of interest.
references
1. Ferrara N, Gerber HP, LeCouter J. The biology of VEGF and its receptors. Nat
Med 2003; 9: 669–673.
2. Llovet JM, Ricci S, Mazzaferro V et al. Sorafenib in advanced hepatocellular
carcinoma. N Engl J Med 2008; 359: 378–390.
3. Motzer RJ, Hutson TE, Tomczak P et al. Sunitinib versus interferon alfa in
metastatic renal-cell carcinoma. N Engl J Med 2007; 356: 115–124.
4. Escudier B, Eisen T, Stadler WM et al. Sorafenib in advanced clear-cell renal-cell
carcinoma. N Engl J Med 2007; 356: 125–134.
5. Sandler A, Gray R, Perry MC et al. Paclitaxel-carboplatin alone or with
bevacizumab for non-small-cell lung cancer. N Engl J Med 2006; 355:
2542–2550.
6. Reck M, von Powel J, Zatloukal P et al. Phase III trial of cisplatin plus gemcitabine
with either placebo or bevacizumab as first-line therapy for nonsquamous non-
small-cell lung cancer: aVAiL. J Clin Oncol 2009; 27: 1227–1234.
7. Ding K, Pater J, Whitehead M et al. Validation of treatment induced specific
adverse effect as a predictor of treatment benefit: a case study of NCIC CTG
BR21. Contemp Clin Trials 2008; 29: 527–536.
8. Wacker B, Nagrani T, Weinberg J et al. Correlation between development of rash
and efficacy in patients treated with the epidermal growth factor receptor
tyrosine kinase inhibitor erlotinib in two large phase III studies. Clin Cancer Res
2007; 13: 3913–3921.
doi:10.1093/annonc/mdq221 | 2225
original article
9. Gatzemeier U, Paz-Ares L, Pereira J et al. Molecular and clinical biomarkers of
cetuximab efficacy: data from the phase III FLEX study in non-small cell lung
cancer (NSCLC). J Thor Oncol 2009; 4 (9 Suppl 1): (abstract B2.3).
10. Jain RK, Duda D, Willett CG et al. Biomarkers of response and resistance to
antiangiogenic therapy. Nat Rev Clin Oncol 2009; 6: 327–338.
11. Sane DC, Anton L, Brosnihan KB. Angiogenic growth factors and hypertension.
Angiogenesis 2004; 7: 193–201.
12. Veronese ML, Mosenkis A, Flaherty KT et al. Mechanisms of hypertension
associated with BAY 43-9006. J Clin Oncol 2006; 24: 1363–1369.
13. Mourad JJ, des Guetz G, Debbabi H, Levy BI. Blood pressure rise following
angiogenesis inhibition by bevacizumab. A crucial role for microcirculation. Ann
Oncol 2008; 19: 927–934.
14. Steeghs N, Gelderblom H, Roodt JO et al. Hypertension and rarefaction during
treatment with telatinib, a small molecule angiogenesis inhibitor. Clin Cancer Res
2008; 14: 3470–3476.
15. Goss G, Arnold A, Shepherd FA et al. A randomized, double-blind trial of
carboplatin and paclitaxel with either daily oral cediranib (AZD 2171) or placebo
in advanced non-small cell lung cancer: NCIC Clinical Trials Group BR24 Study.
J Clin Oncol 2010; 28: 49–55.
16. Therasse P, Arbuck SG, Eisenhauer EA et al. New guidelines to evaluate the
response to treatment in solid tumors. J Natl Cancer Inst 2000; 92: 205–216.
17. Laurie SA, Gauthier I, Arnold A et al. Phase I and pharmacokinetic study of daily
oral AZD2171, an inhibitor of vascular endothelial growth factor tyrosine kinases,
in combination with carboplatin and paclitaxel in patients with advanced non-
small-cell lung cancer: the National Cancer Institute of Canada Clinical Trials
Group. J Clin Oncol 2008; 26: 1871–1878.
18. Curwen JO, Musgrove HL, Kendrew J et al. Inhibition of vascular endothelial
growth factor—A signaling induces hypertension: examining the effect of
cediranib (Recentin; AZD2171) treatment on blood pressure in rat and the use of
concomitant antihypertensive therapy. Clin Cancer Res 2008; 14: 3124–3131.
19. Zhu X, Wu S, Dahut WL, Parikh CR. Risks of proteinuria and hypertension with
bevacizumab, an antibody against vascular endothelial growth factor: systematic
review and meta-analysis. Am J Kidney Dis 2007; 49: 186–193.
20. Wu S, Chen JJ, Kudelka A et al. Incidence and risk of hypertension with
sorafenib in patients with cancer: a systematic review and meta-analysis. Lancet
Oncol 2008; 9: 86–87.
21. Zhu X, Stergiopoulos K, Wu S. Risk of hypertension and renal dysfunction with an
angiogenesis inhibitor sunitinib: systematic review and meta-analysis. Acta Oncol
2009; 48: 9–17.
22. Azizi M, Chedid A, Oudard S. Home blood-pressure monitoring in patients
receiving sunitinib. N Engl J Med 2008; 358: 95–97.
23. Yang JC, Haworth L, Sherry RM et al. A randomized trial of bevacizumab, and
anti-vascular endothelial growth factor antibody, for metastatic renal cancer.
N Engl J Med 2003; 349: 427–434.
24. Johnson DH, Fehrenbacher L, Novotny WF et al. Randomized phase II trial
comparing bevacizumab plus carboplatin and paclitaxel with carboplatin and
paclitaxel alone in previously untreated locally advanced or metastatic non-small-
cell lung cancer. J Clin Oncol 2004; 22: 2184–2191.
25. Drevs J, Siegert P, Medinger M et al. Phase I clinical study of AZD2171, an
oral vascular endothelial growth factor signaling inhibitor, in patients with
advanced solid tumors. J Clin Oncol 2007; 25: 3045–3054.
2226 | Goodwin et al.
Annals of Oncology
26. Goss G, Shepherd FA, Laurie S et al. A phase I and pharmacokinetic study of
daily oral cediranib, an inhibitor of vascular endothelial growth factor tyrosine
kinases, in combination with cisplatin and gemcitabine in patients with advanced
non-small cell lung cancer: a study of the National Cancer Institute of Canada
Clinical Trials Group. Eur J Cancer 2009; 45: 782–788.
27. Dahlberg SE, Sandler A, Brahmer J et al. Clinical course of advanced non-small
cell lung cancer (NSCLC) patients (pts) experiencing hypertension (HTN)
during treatment (TX) with bevacizumab (B) in combination with carboplatin
(C) and paclitaxel (P) on E4599. J Clin Oncol 2009; 27 (15s Suppl): (abstract
8042).
28. Leighl NB, Paz-Ares L, Douillard JY et al. Randomized phase III study of matrix
metalloproteinase inhibitor BMS-275291 in combination with paclitaxel and
carboplatin in advanced non-small-cell lung cancer: National Cancer Institute
of Canada-Clinical Trials Group study BR.18. J Clin Oncol 2005; 23:
2831–2839.
29. Kabbinavar F, Schulz J, McCleod M et al. Addition of bevacizumab to bolus
fluorouracil and leucovorin in first-line metastatic colorectal cancer: results of
a randomized phase II trial. J Clin Oncol 2005; 23: 3697–3705.
30. Hurwitz H, Ferhenbacher L, Novotny W et al. Bevacizumab plus irinotecan,
fluorouracil, and leucovorin for metastatic colorectal cancer. N Engl J Med 2004;
350: 2335–2342.
31. Van Cutsem E, Vervenne WL, Bennouna J et al. Phase III trial of bevacizumab in
combination with gemcitabine and erlotinib in patients with metastatic pancreatic
cancer. J Clin Oncol 2009; 27: 2231–2237.
32. Scagliotti G, von Powel J, Reck M et al. Phase III trial comparing carboplatin and
paclitaxel with or without sorafenib in chemonaive patients with stage IIIB (with
effusion) or IV non-small cell lung cancer. Presented at IASLC-ESMO 1st
European Lung Cancer Conference; April 2008, Geneva, Switzerland.
33. Scartozzi M, Galizia E, Chiorinni S et al. Arterial hypertension correlates with
clinical outcome in colorectal cancer. patients treated with first-line
bevacizumab. Ann Oncol 2009; 20: 227–230.
34. Friberg G, Kasza K, Vokes EE, Kindler HL. Early hypertension (HTN) as a potential
pharmacodynamic (PD) marker for survival in pancreatic cancer (PC) patients
(pts) treated with bevacizumab (B) and gemcitabine (G). J Clin Oncol 2005; 23
(16s Suppl): (abstract 3020).
35. Rixe O, Billemont B, Izzedine H. Hypertension as a predictive factor of sunitinib
activity. Ann Oncol 2007; 18: 1117.
36. Rixe O, Dutcher J, Motzer R et al. Diastolic blood pressure (dBP) and
pharmacokinetics (PK) as predictors of axitinib efficacy in metastatic renal cell
cancer (mRCC). J Clin Oncol 2009; 27 (15s Suppl): (abstract 5045).
37. Thatcher N, Tsai CM, Laskin J et al. MO19390 (SAiL): hypertension in patients
with advanced of recurrent non-squamous non-small cell lung cancer (NSCLC)
receiving first-line bevacizumab plus chemotherapy. J Thor Oncol 2009; 4
(Suppl): (abstract C2.4).
38. Belotti D, Vergani V, Drudis T et al. The microtubule-affecting drug paclitaxel has
antiangiogenic activity. Clin Cancer Res 1996; 2: 1843–1849.
39. Schneider BP, Wang M, Radovich M et al. Association of vascular endothelial
growth factor and vascular endothelial growth factor receptor-2 genetic
polymorphisms with outcome in a trial of paclitaxel compared with paclitaxel plus
bevacizumab in advanced breast cancer: ECOG 2100. J Clin Oncol 2008; 26:
4672–4678.
Volume 21 | No. 11 | November 2010