Improved Survival with Bevacizumab in Advanced Cervical Cancer
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
The n e w e ng l a n d j o u r na l of m e dic i n e
original article
Improved Survival with Bevacizumab
in Advanced Cervical Cancer
Krishnansu S. Tewari, M.D., Michael W. Sill, Ph.D., Harry J. Long III, M.D.,
Richard T. Penson, M.D., Helen Huang, M.S., Lois M. Ramondetta, M.D.,
Lisa M. Landrum, M.D., Ana Oaknin, M.D., Thomas J. Reid, M.D.,
Mario M. Leitao, M.D., Helen E. Michael, M.D., and Bradley J. Monk, M.D.
A BS T R AC T
Background
From the University of California, Irvine, Vascular endothelial growth factor (VEGF) promotes angiogenesis, a mediator of
Medical Center, Orange (K.S.T.); Roswell disease progression in cervical cancer. Bevacizumab, a humanized anti-VEGF mono
Park Cancer Institute, State University of
New York at Buffalo, Buffalo (M.W.S., H.H.); clonal antibody, has single-agent activity in previously treated, recurrent disease.
Mayo Clinic, Rochester, MN (H.J.L.); Massa- Most patients in whom recurrent cervical cancer develops have previously received
chusetts General Hospital, Boston (R.T.P.); cisplatin with radiation therapy, which reduces the effectiveness of cisplatin at the
M.D. Anderson Cancer Center, Houston
(L.M.R.); University of Oklahoma, Oklaho- time of recurrence. We evaluated the effectiveness of bevacizumab and nonplatinum
ma City (L.M.L.); Vall d’Hebron University combination chemotherapy in patients with recurrent, persistent, or metastatic cer-
Hospital, Barcelona (A.O.); University of vical cancer.
Cincinnati College of Medicine–Women’s
Cancer Center at Kettering, Kettering, OH Methods
(T.J.R.); Memorial Sloan-Kettering Can-
cer Center, New York (M.M.L.); Indiana
Using a 2-by-2 factorial design, we randomly assigned 452 patients to chemotherapy
University School of Medicine, Indianapo- with or without bevacizumab at a dose of 15 mg per kilogram of body weight. Chemo-
lis (H.E.M.); and the University of Arizona therapy consisted of cisplatin at a dose of 50 mg per square meter of body-surface area,
Cancer Center and Creighton University
at St. Joseph’s Hospital and Medical
plus paclitaxel at a dose of 135 or 175 mg per square meter or topotecan at a dose of
Center, Phoenix (B.J.M.). Address reprint 0.75 mg per square meter on days 1 to 3, plus paclitaxel at a dose of 175 mg per square
requests to Dr. Tewari at the Division meter on day 1. Cycles were repeated every 21 days until disease progression, the de-
of Gynecologic Oncology, University of
California, Irvine, Medical Center, 101
velopment of unacceptable toxic effects, or a complete response was documented. The
City Dr. S., Bldg. 56, Orange, CA 92868, primary end point was overall survival; a reduction of 30% in the hazard ratio for death
or at ktewari@uci.edu. was considered clinically important.
N Engl J Med 2014;370:734-43.
Results
DOI: 10.1056/NEJMoa1309748
Copyright © 2014 Massachusetts Medical Society Groups were well balanced with respect to age, histologic findings, performance sta-
tus, previous use or nonuse of a radiosensitizing platinum agent, and disease status.
Topotecan–paclitaxel was not superior to cisplatin–paclitaxel (hazard ratio for death,
1.20). With the data for the two chemotherapy regimens combined, the addition
of bevacizumab to chemotherapy was associated with increased overall survival
(17.0 months vs. 13.3 months; hazard ratio for death, 0.71; 98% confidence interval,
0.54 to 0.95; P = 0.004 in a one-sided test) and higher response rates (48% vs. 36%,
P = 0.008). Bevacizumab, as compared with chemotherapy alone, was associated with
an increased incidence of hypertension of grade 2 or higher (25% vs. 2%), thrombo
embolic events of grade 3 or higher (8% vs. 1%), and gastrointestinal fistulas of
grade 3 or higher (3% vs. 0%).
Conclusions
The addition of bevacizumab to combination chemotherapy in patients with recur-
rent, persistent, or metastatic cervical cancer was associated with an improvement
of 3.7 months in median overall survival. (Funded by the National Cancer Institute;
GOG 240 ClinicalTrials.gov number, NCT00803062.)
734 n engl j med 370;8 nejm.org february 20, 2014
The New England Journal of Medicine
Downloaded from nejm.org at DIGNITY HEALTH on February 19, 2014. For personal use only. No other uses without permission.
Copyright © 2014 Massachusetts Medical Society. All rights reserved.Bevacizumab in Advanced Cervical Cancer
R
ates of cervical cancer in devel- they were candidates for curative therapy by means
oped countries have decreased dramati- of pelvic exenteration. All cancers were confirmed
cally because of cytologic screening and by a central pathology laboratory. A GOG perfor-
DNA testing for high-risk human papillomavirus mance status score of 0 or 1 (on a scale of 0 to 4,
(HPV) types. Approximately 12,000 cases of cer- with 0 indicating that the person is fully active
vical cancer are diagnosed in the United States and 1 indicating that the person is restricted in
annually, and with continued increases in HPV physically strenuous activities but ambulatory)
vaccination, numbers of cases are expected to de- was required, and patients had to have adequate
crease further.1 However, for vulnerable popula- renal, hepatic, and bone marrow function. All pa-
tions without access to health care in the United tients were required to have measurable disease.
States and throughout the world, cervical cancer Patients treated with chemotherapy for recurrence
remains a considerable problem, with 500,000 new and those with nonhealing wounds, active bleeding
cases and 250,000 deaths annually.2 Although conditions, or inadequately anticoagulated throm-
early-stage and locally advanced cancers may be boembolism were ineligible. All patients provided
cured with radical surgery, chemoradiotherapy, written informed consent before enrollment.
or both, patients with metastatic cancers and
those with persistent or recurrent disease after Study Design and Treatment
platinum-based chemoradiotherapy have limited Patients were randomly assigned to one of four
options.3-17 Nonplatinum combination chemo- intravenous regimens that were repeated at 21-day
therapy has been proposed as a strategy to cir- intervals. Control treatment consisted of cisplatin
cumvent platinum resistance, but new forms of (at a dose of 50 mg per square meter of body-
therapy are needed. surface area) plus paclitaxel (at a dose of 135 or
Vascular endothelial growth factor (VEGF) is 175 mg per square meter on day 1). The non-
a key mediator of tumor angiogenesis, a process platinum combination chemotherapy consisted
that correlates directly with the extent of disease of topotecan (at a dose of 0.75 mg per square
and inversely with survival.18 Bevacizumab, a hu- meter on days 1 to 3) plus paclitaxel (at a dose of
manized VEGF-neutralizing monoclonal antibody, 175 mg per square meter on day 1). Each of these
has single-agent activity in heavily pretreated, regimens was studied with and without bevaciz
recurrent cervical carcinoma.19,20 In GOG 240, umab (at a dose of 15 mg per kilogram of body
a phase 3, randomized trial performed in the weight on day 1). Treatment was discontinued at
United States and in Spain through the Gyneco- the onset of d
isease progression or the develop-
logic Oncology Group (GOG) and the Spanish ment of unacceptable toxic effects, or if the pa-
Research Group for Ovarian Cancer, we investi- tient had a complete response.
gated the incorporation of bevacizumab and the
use of nonplatinum combination chemotherapy Assessments
in the treatment of advanced cervical cancer. Disease was assessed by means of physical ex-
amination and chest radiography, as well as by
Me thods means of computed tomography or magnetic
resonance imaging of the abdomen and pelvis
Study Oversight within 28 days before the study treatment was
The study was sponsored by the National Cancer initiated. In patients without disease progres-
Institute, which provided bevacizumab without sion, imaging was repeated every other cycle.
charge. All the authors wrote the manuscript and Tumor measurements according to the Response
take responsibility for the accuracy and com- Evaluation Criteria in Solid Tumors (RECIST),
pleteness of the reported data and for the fidelity version 1, were made within 1 week before the
of the study to the protocol, which is available next planned cycle.21 After discontinuation of
with the full text of this article at NEJM.org. treatment, disease was assessed every 3 months
for 2 years, followed by assessment every
Patients 6 months for 3 years until disease progression
Patients with metastatic, persistent, or recurrent was documented.
cervical carcinoma were eligible for the study. Pa- Three validated, sensitive instruments were used
tients with recurrent disease were excluded if to measure health-related quality of life. The Trial
n engl j med 370;8 nejm.org february 20, 2014 735
The New England Journal of Medicine
Downloaded from nejm.org at DIGNITY HEALTH on February 19, 2014. For personal use only. No other uses without permission.
Copyright © 2014 Massachusetts Medical Society. All rights reserved.The n e w e ng l a n d j o u r na l of m e dic i n e
Outcome Index of the Functional Assessment of tizing platinum, and disease status (recurrence
Cancer Therapy (FACT)–Cervix (FACT-Cx-TOI) sur- or persistence of disease vs. advanced primary
vey was used to assess physical and functional disease).
well-being (on a scale from 0 to 4, with higher The primary end points were overall survival
scores indicating worsening well-being). Pain was and the frequency and severity of adverse events
measured with the use of the Brief Pain Inven- associated with each regimen. Progression-free
tory (BPI) (on a scale from 0 to 10, with higher survival and the response rate were secondary end
scores indicating more severe pain). Neurotoxicity points. Differences in overall survival and pro-
was measured with the use of the neurotoxicity gression-free survival according to intervention
subscale short form (FACT/GOG-NTX) (on a scale level were assessed primarily by means of the
from 0 to 4, with higher scores indicating in- log-rank test, stratified according to clinical
creased neurotoxicity).22 Baseline assessments were prognostic markers and the level of the other in-
completed before randomization, before cycles 2 tervention.26 Hazard ratios were estimated with
and 5, and 6 and 9 months after cycle 1. the use of a Cox proportional-hazards model.27
Safety, as assessed according to the National We calculated that we would need to enroll
Cancer Institute Common Terminology Criteria approximately 450 patients, with approximately
for Adverse Events, was monitored during each 346 deaths expected, to provide the study with
cycle.23,24 Myeloid growth factor was permitted 90% power to detect a reduction in the risk of
only for hospitalized patients with grade 3 or death of at least 30% with either experimental
higher febrile neutropenia (absolute neutrophil treatment, with the one-sided type I error rate
count, 150 mm Hg health-related quality of life were evaluated with
or diastolic blood pressure >100 mm Hg), protein- the use of a mixed model for analysis of re-
uria (urine protein-to-creatinine ratio ≥3.5), arte- peated measures.30
rial thrombosis, venous thrombosis, coagulopathy,
or intestinal obstruction or disruption. R e sult s
Statistical Analysis Patients
The statistical analysis plan is available with the Between April 2009 and January 2012, a total of
protocol at NEJM.org. Assuming an absence of 452 women were enrolled from 164 institutions
interaction between experimental agents, we used in the United States and Spain. The data freezes
a 2-by-2 factorial design to investigate the effect of occurred on February 6, 2012, and December 12,
anti-VEGF therapy (bevacizumab) and a regi- 2012. Analyses provided in this article concerning
men of nonplatinum combination chemotherapy the bevacizumab regimens are from the second
(topotecan–paclitaxel).25 The study was based on data freeze. Figure 1 shows randomization and
the intention-to-treat principle. Patients were pro- follow-up among patients assigned to chemother-
spectively stratified according to GOG perfor- apy with or without bevacizumab.
mance status, prior use or nonuse of radiosensi- Demographic characteristics and clinical and
736 n engl j med 370;8 nejm.org february 20, 2014
The New England Journal of Medicine
Downloaded from nejm.org at DIGNITY HEALTH on February 19, 2014. For personal use only. No other uses without permission.
Copyright © 2014 Massachusetts Medical Society. All rights reserved.Bevacizumab in Advanced Cervical Cancer
pathological factors were evenly distributed be- 30), and for those who received chemotherapy plus
tween the treatment groups (see Table S15 in bevacizumab, the median was 7 (range, 0 to 36).
the Supplementary Appendix). The majority of Ninety-seven percent of patients discontinued
patients (72%) had recurrent disease, and 11% the study treatment; the most common reason
of patients had persistent disease. More than was disease progression (in 51% of patients
70% of patients in each group had previously who received chemotherapy [either regimen]
received platinum-based chemoradiotherapy. alone and 38% of patients who received chemo-
The median number of cycles for patients therapy [either regimen] plus bevacizumab).
treated with chemotherapy alone was 6 (range, 0 to Treatment was discontinued owing to adverse
425 Patients were enrolled and underwent
randomization
114 Were assigned to receive 111 Were assigned to receive 115 Were assigned to receive 112 Were assigned to receive
cisplatin (50 mg/m2) plus topotecan (0.75 mg/m2, days 1–3) cisplatin (50 mg/m2) plus pacli- topotecan (0.75 mg/m2, days 1–3)
paclitaxel (135 or 175 mg/m2) plus paclitaxel (175 mg/m2) taxel (135 or 175 mg/m2) plus plus paclitaxel (175 mg/m2) plus
bevacizumab (15 mg/kg) bevacizumab (15 mg/kg)
225 Were assigned to chemotherapy alone 227 Were assigned to chemotherapy plus
with cycles repeated every 21 days bevacizumab with cycles repeated
every 21 days
219 Were included in safety analysis 220 Were included in safety analysis
6 Did not have adverse-event data submitted 7 Did not have adverse-event data submitted
225 Were included in efficacy analysis 227 Were included in efficacy analysis
184 Reached PFS end point 183 Reached PFS end point
140 Died 131 Died
225 Discontinued treatment 51 Crossed over to salvage therapy 227 Discontinued treatment 33 Crossed over to salvage
6 Had complete response to 12 to bevacizumab 15 Had complete response to therapy
assigned treatment 20 to cisplatin assigned treatment 7 to bevacizumab
115 Had disease progression 16 to topotecan 86 Had disease progression 12 to cisplatin
36 Had toxic effects 1 to both bevacizumab and 57 Had toxic effects 13 to topotecan
35 Declined further treatment topotecan 28 Declined further treatment 0 to both bevacizumab and
5 Died 2 to both bevacizumab 6 Died topotecan
2 Had other disease and cisplatin 4 Had other disease 1 to both bevacizumab
21 Had other reasons 23 Had other reasons and cisplatin
5 Had unspecified reason 8 Had unspecified reason
Figure 1. Enrollment, Randomization, and Follow-up of the Study Patients.
All 452 patients (225 in the chemotherapy-alone group and 227 in the chemotherapy-plus-bevacizumab group) who underwent randomization
were included in the efficacy analysis. Because data on adverse events were not included for 6 patients in the chemotherapy-alone group, the
safety analysis in this group included 219 patients. Because data on adverse events were not included for 7 patients in the chemotherapy-plus-
bevacizumab group, the safety analysis in this group included 220 patients. A total of 15 patients randomly assigned to chemotherapy alone
crossed over to salvage bevacizumab at the time of disease progression. PFS denotes progression-free survival.
n engl j med 370;8 nejm.org february 20, 2014 737
The New England Journal of Medicine
Downloaded from nejm.org at DIGNITY HEALTH on February 19, 2014. For personal use only. No other uses without permission.
Copyright © 2014 Massachusetts Medical Society. All rights reserved.The n e w e ng l a n d j o u r na l of m e dic i n e
events in a higher percentage of patients in the Treatment with cisplatin–paclitaxel–bevacizu
chemotherapy–bevacizumab group than in the mab, as compared with cisplatin–paclitaxel alone,
chemotherapy-alone group (25% vs. 16%). was associated with a hazard ratio for death of
0.68 (95% CI, 0.48 to 0.97) (Fig. 3C). The response
Efficacy rates were 50% (cisplatin–paclitaxel–bevacizumab)
At the time of the interim analysis (the first data and 45% (cisplatin–paclitaxel) (P = 0.51, two-sided
freeze), 62% of patients were alive, with a median test); 17 patients and 9 patients had a complete re-
follow-up of 12.5 months. As compared with cis- sponse, respectively. Topotecan–paclitaxel–bevaciz
platin–paclitaxel (either with or without bevaciz umab, as compared with topotecan–paclitaxel
umab) topotecan–paclitaxel was associated with a alone, was associated with a hazard ratio for
significantly higher risk of progression (hazard death of 0.74 (95% CI, 0.53 to 1.05) (Fig. 3D). The
ratio, 1.39; 95% confidence interval [CI], 1.09 to response rates were 47% (topotecan–paclitaxel–
1.77) (Fig. 2A), but it did not significantly affect bevacizumab) and 27% (topotecan–paclitaxel)
overall survival (hazard ratio for death, 1.20; 99% (P = 0.002, two-sided test); 11 patients and 5 pa-
CI, 0.82 to 1.76) (Fig. 2B). There was also no sig- tients had a complete response, respectively.
nificant difference in mortality between the che- Figure 3E shows multiple prognostic factors.
motherapy regimens in the subgroup of patients The treatment benefit with bevacizumab was also
with previous exposure to platinum (hazard ratio, observed in subgroup analyses of age, perfor-
1.18; 95% CI, 0.84 to 1.65) and in the subgroup mance status, race, squamous histologic type,
with no previous exposure to platinum (hazard status with respect to prior platinum exposure,
ratio, 1.35; 95% CI, 0.68 to 2.69) (Fig. S3 and S4 recurrent or persistent disease, and pelvic loca-
in the Supplementary Appendix). Noting that tion of the target lesion.
topotecan was not a superior (or an inferior) sub-
stitute for cisplatin, the data and safety monitor- Quality of Life
ing committee voted on March 12, 2012, for early The rate of compliance with health-related qual-
release of data from this first data freeze to all ity of life surveys was 96%, 84%, 78%, 67%, and
investigators and patients. 63% among patients at cycles 1, 2, and 5, and at
At a median follow-up of 20.8 months, 6 months and 9 months of follow-up, respectively,
271 deaths had been reported (60% of the to- and was balanced between treatment groups
tal study population), and the results from the (P = 0.67). The mean FACT-Cx-TOI scores exceeded
second data freeze were partially released after a 70 at each time point in each group. The fitted
recommendation by the data and safety moni- mixed-model estimates for the FACT-Cx-TOI and
toring committee. The interaction term was not BPI scores indicated that the addition of bevaciz
significant, indicating that there was no interaction umab did not adversely affect health-related
between the two treatment regimens under investi- quality of life; scores for patients who received
gation. The incorporation of bevacizumab signif antiangiogenic therapy were 1.2 points lower, on
icantly improved the median overall survival as average than scores for patients who did not re-
compared with chemotherapy alone (17.0 months ceive such therapy, although the difference was
vs. 13.3 months; hazard ratio for death, 0.71; not significant (99% CI, −4.1 to 1.7; P = 0.30). The
98% CI, 0.54 to 0.95) (Fig. 3A). A significant im- fitted mixed-effects mixed-distribution model es-
provement in progression-free survival was also timates for the FACT/GOG-NTX scores showed a
seen (8.2 vs. 5.9 months; hazard ratio for disease nonsignificant trend for patients receiving beva
progression, 0.67; 95% CI, 0.54 to 0.82) (Fig. 3B). cizumab to report fewer neurotoxic symptoms
The response rate was significantly higher among (overall odds ratio, 0.58; 99% CI, 0.29 to 1.17;
patients who received bevacizumab than among P = 0.05), and the severity of neurotoxic symptoms
those who did not receive bevacizumab (48% vs. reported was similar in the two groups (P = 0.70).
36%) (relative probability of a response, 1.35;
95% CI, 1.08 to 1.68; P = 0.008, two-sided test). Safety
Among patients who received bevacizumab, 28 had Table 1 shows the frequency of adverse events
a complete response, and among those who re- potentially associated with bevacizumab. Hyper-
ceived chemotherapy alone, 14 had a complete tension of grade 2 or higher was significantly
response (P = 0.03). Treatment was discontinued more common with bevacizumab-containing
in 21 patients who had a complete response. regimens than with regimens that did not con-
738 n engl j med 370;8 nejm.org february 20, 2014
The New England Journal of Medicine
Downloaded from nejm.org at DIGNITY HEALTH on February 19, 2014. For personal use only. No other uses without permission.
Copyright © 2014 Massachusetts Medical Society. All rights reserved.Bevacizumab in Advanced Cervical Cancer tain bevacizumab (25% vs. 2%, P
740
A Median
B Median C Median
Events Overall Progression-free Overall
no. (%) Survival Events Survival Events Survival
mo no. (%) mo no. (%) mo
Chemotherapy (N=225) 140 (62) 13.3 Chemotherapy (N=225) 184 (82) 5.9 CP (N=114) 69 (61) 14.3
Chemotherapy+Bev (N=227) 131 (58) 17.0 Chemotherapy+Bev (N=227) 183 (81) 8.2 CP+Bev (N=115) 66 (58) 17.5
Hazard ratio, 0.71 (98% CI, 0.54–0.95); Hazard ratio, 0.67 (95% CI, 0.54–0.82); two-sided P=0.002 Hazard ratio, 0.68 (95% CI, 0.48–0.97); one-sided P=0.04
1.0 one-sided P=0.004 1.0 1.0
Median follow-up, 20.8 mo
0.8 0.8 0.8
0.6 0.6 0.6
0.4 0.4 0.4
Probability of
0.2 0.2 0.2
Probability of Survival
Probability of Survival
Progression-free Survival
0.0 0.0 0.0
0 6 12 18 24 30 36 0 6 12 18 24 30 36 0 6 12 18 24 30 36
Months since Randomization Months since Randomization Months since Randomization
The
No. at Risk No. at Risk No. at Risk
Chemotherapy 225 167 94 45 17 8 Chemotherapy 225 103 40 14 6 3 CP 114 89 50 22 12 5
Chemotherapy 227 184 121 69 30 10 Chemotherapy 227 132 70 22 6 3 CP+bev 115 94 63 37 17 5
+bev +bev
D Median
E
Overall
n engl j med 370;8
Events Survival Subgroup No. of Patients Hazard Ratio
no. (%) mo
Age ≤40 yr 112
TP (N=111) 71 (64) 12.7 40 to ≤48 yr 111
TP+Bev (N=112) 65 (58) 16.2 48 to ≤56 yr 108
nejm.org
Hazard ratio, 0.74 (95% CI, 0.53–1.05); one-sided P=0.09 >56 yr 121
Performance status 0 263
n e w e ng l a n d j o u r na l
1.0 1 189
The New England Journal of Medicine
Previous platinum No
of
115
0.8 radiation therapy
Yes 337
0.6 Disease status Advanced 76
Recurrent or persistent 376
0.4
february 20, 2014
Topotecan treatment No 229
Yes 223
Copyright © 2014 Massachusetts Medical Society. All rights reserved.
0.2
m e dic i n e
Race Not black 392
Probability of Survival
0.0 Black 60
0 6 12 18 24 30 36 Histologic type Adenocarcinoma 86
Adenosquamous 44
Months since Randomization Other 12
No. at Risk Squamous 310
TP 111 78 44 23 5 3 Pelvic disease No 210
TP+bev 115 90 58 32 13 5 Yes 242
Overall 452
0.0 0.5 1.0 1.5 2.0 2.5
Experimental Better Control Better
Downloaded from nejm.org at DIGNITY HEALTH on February 19, 2014. For personal use only. No other uses without permission.Bevacizumab in Advanced Cervical Cancer
maps to the C-terminal and correlates with
Figure 3 (facing page). Effect of Incorporation
of Bevacizumab on Survival. displacement by E7 of the histone deacetylases
Panel A shows overall survival and Panel B shows pro- HDAC1, HDAC4, and HDAC7.37
gression-free survival among patients who received ei- Short-lived responses to chemotherapy in pa-
ther chemotherapy regimen plus bevacizumab or either tients with advanced cervical cancer indicate
chemotherapy regimen alone. Panel C shows overall that the disease is relatively chemorefractory. We
survival among patients who received cisplatin–paclitaxel
selected overall survival as the primary end point
with or without bevacizumab, and Panel D shows overall
survival among those who received topotecan–paclitaxel because, unlike patients with other cancers, pa-
with or without bevacizumab. In Panel E, a forest plot tients with advanced cervical cancer usually do
shows the effect of chemotherapy with bevacizumab not have a sustained response to chemotherapy
(experimental), as compared with chemotherapy with- and cannot receive multiple lines of chemother-
out bevacizumab (control), on overall survival, stratified
apy because of unacceptable side effects. We
according to multiple prognostic factors. A positive treat-
ment benefit (orange) is indicated. The upper bound of think that the 3.7-month improvement in me-
the confidence interval for histologic type was truncated dian overall survival attributed to the addition
at 2.5. Its true upper bound, 7.07, resulted from a small of bevacizumab to chemotherapy is clinically
sample. An interaction test was performed, and nothing meaningful. Antiangiogenic therapy and possi-
was found to be significant. The factor with the smallest
bly other targeted agents may provide additional
P value was histologic type (P = 0.07).
gains in survival time, allowing for multiple
lines of therapy with sustained health-related
quality of life. Given the well-recognized HPV
an aggressive course in cervical cancer. Vascular epidemic, these data provide support for further
markings seen at colposcopy in women with ab- investigation of antivascular therapy in patients
normal Papanicolaou tests are hallmarks for in- with other HPV-induced tumors, including vulvar,
vasive disease, and increased microvessel density anal, penile, and oropharyngeal carcinomas.
and strong immunostaining for the endothelial- Two additional agents that may have activity in
cell marker, CD31, in cervical cancers suggest a advanced cervical cancer are pazopanib, an intra-
poor prognosis.2 VEGF is involved in mitogenesis, cellular small-molecule tyrosine kinase inhib
angiogenesis, endothelial-cell survival, and induc- itor that targets VEGF receptor, and sorafenib,
tion of hematopoiesis.33 Patients with high-grade a multikinase inhibitor.38 Data are lacking on
cervical dysplasia and invasive carcinoma have in- drugs that inhibit angiogenesis through non–
creased expression of VEGF and hypoxia-inducible VEGF-dependent pathways (e.g., the Tie2–angio-
factor 1α (HIF-1α).34 The invasive phenotype is poietin-2 pathway), as well as vascular disrupting
present only with up-regulated VEGF. Overexpres- agents (e.g., vadimezan). Finally, drugs targeting
sion of oncogenic HPV subtypes enhances HIF-1α nonangiogenic signal-transduction pathways that
protein accumulation and VEGF expression. are integral to tumor progression may be con-
The molecular mechanism through which sidered, including Wee1 checkpoint inhibitors
HPV mediates tumor angiogenesis has been elu- and Notch γ-secretase inhibitors, the latter be-
cidated. In the native form, HPV exists as circular ing an evolutionarily conserved cell-fate deci-
double-stranded DNA episomes, and viral E2 ex- sion switch in cervical cancer.
pression prevents transcription of the viral onco- There has been a large unmet medical need
genes, E6 and E7. The E2 reading frame is dis- for active treatments for cervical cancer, which
rupted on viral integration into host DNA, is a leading cause of death from cancer in develop-
resulting in lack of repression of E6 and E7, which ing countries. In the poorest regions, where rates
mediate neoplastic transformation through degra- are highest (sub-Saharan Africa, Latin America,
dation or inactivation of cellular tumor-suppressor and Southeast Asia, including India), many wom-
protein p53 and retinoblastoma protein, respec- en are forced by socioeconomic and political cir-
tively.35 VEGF isoform expression can be con- cumstances to act as the sole provider for their
siderably reduced by silencing HPV E6 messen- young families. With their deaths, the effect on
ger RNA with specific small interfering RNAs families can be devastating. The improvement in
but not when p53 is silenced, suggesting that survival that is conferred by cisplatin–paclitaxel–
E6 induces VEGF through a p53-independent bevacizumab treatment warrants cost-effectiveness
mechanism.36 HIF-1α activity enhanced by E7 studies because of the societal burden involved in
n engl j med 370;8 nejm.org february 20, 2014 741
The New England Journal of Medicine
Downloaded from nejm.org at DIGNITY HEALTH on February 19, 2014. For personal use only. No other uses without permission.
Copyright © 2014 Massachusetts Medical Society. All rights reserved.The n e w e ng l a n d j o u r na l of m e dic i n e
Table 1. Selected Adverse Events among the Study Patients, According to Treatment Group.*
Chemotherapy Chemotherapy plus
Alone Bevacizumab Odds Ratio
Event (N = 219) (N = 220) (95% CI) P Value
no. of patients (%)
Gastrointestinal events, excluding 96 (44) 114 (52) 1.38 (0.93–2.04) 0.10
fistulas (grade ≥2)
Fistula (grade ≥3)
Gastrointestinal 0 7 (3) NA (1.90–∞) 0.02
Genitourinary 1 (Bevacizumab in Advanced Cervical Cancer
of carboplatin and iproplatin in advanced ogy Group study. J Clin Oncol 2009;27: 31. Bahadori HR, Green MR, Catapano
squamous carcinoma of the uterine cer- 1069-74. CV. Synergistic interaction between topot
vix: a Gynecologic Oncology Group study. 20. Ferrara N, Hillan KJ, Gerber HP, ecan and microtubule-interfering agents.
J Clin Oncol 1989;7:1462-8. Novotny W. Discovery and development of Cancer Chemother Pharmacol 2001;48:
12. Omura GA, Blessing JA, Vaccarello L, bevacizumab, an anti-VEGF antibody for 188-96.
et al. Randomized trial of cisplatin versus treating cancer. Nat Rev Drug Discov 32. Tiersten AD, Selleck MJ, Hershman
cisplatin plus mitolactol versus cisplatin 2004;3:391-400. DL, et al. Phase II study of topotecan and
plus ifosfamide in advanced squamous 21. Therasse P, Arbuck SG, Eisenhauer paclitaxel for recurrent, persistent, or
carcinoma of the cervix: a Gynecologic EA, et al. New guidelines to evaluate the metastatic cervical carcinoma. Gynecol
Oncology Group study. J Clin Oncol 1997; response to treatment in solid tumors: Oncol 2004;92:635-8.
15:165-71. European Organization for Research and 33. No JH, Jo H, Kim SH, et al. Expression
13. Bloss JD, Blessing JA, Behrens BC, et Treatment of Cancer, National Cancer In- of vascular endothelial growth factor and
al. Randomized trial of cisplatin and ifos- stitute of Canada. J Natl Cancer Inst hypoxia inducible factor-1alpha in cervi-
famide with or without bleomycin in 2000;92:205-16. cal neoplasia. Ann N Y Acad Sci 2009;
squamous carcinoma of the cervix: a Gy- 22. Cella DF. Manual for the Functional 1171:105-10.
necologic Oncology Group study. J Clin Assessment of Cancer Therapy (FACT) 34. Tang X, Zhang Q, Nishitani J, Brown
Oncol 2002;20:1832-7. measurement system (version 4). Chicago: J, Shi S, Le AD. Overexpression of human
14. Moore DH, Blessing JA, McQuellon RP, Northwestern University, Center for Out- papillomavirus type 16 oncoproteins en-
et al. Phase III study of cisplatin with or comes, Research and Education (CORE), hances hypoxia-inducible factor 1 alpha
without paclitaxel in stage IVB, recurrent, 1997. protein accumulation and vascular endo-
or persistent squamous cell carcinoma of 23. Common Terminology Criteria for Ad- thelial growth factor expression in hu-
the cervix: a Gynecologic Oncology Group verse Events (CTCAE), v3.0. Bethesda, MD: man cervical carcinoma cells. Clin Cancer
study. J Clin Oncol 2004;22:3113-9. Cancer Therapy Evaluation Program, 2006 Res 2007;13:2568-76.
15. Long HJ III, Bundy BN, Grendys EC Jr, (http://ctep.cancer.gov/protocolDevelopment/ 35. Tewari KS, Taylor JA, Liao SY, et al.
et al. Randomized phase III trial of cispla- electronic_applications/ctc.htm#ctc_30). Development and assessment of a gener-
tin with or without topotecan in carcino- 24. Common Terminology Criteria for Ad al theory of cervical carcinogenesis uti-
ma of the uterine cervix: a Gynecologic verse Events (CTCAE). v4.0. Bethesda, MD: lizing a severe combined immunodefi-
Oncology Group study. J Clin Oncol 2005; Cancer Therapy Evaluation Program, 2011 ciency murine-human xenograft model.
23:4626-33. (http://ctep.cancer.gov/protocolDevelopment/ Gynecol Oncol 2000;77:137-48.
16. Moore DH, Tian C, Monk BJ, Long HJ, electronic_applications/ctc.htm#ctc_40). 36. Clere N, Bermont L, Fauconnet S, et
Omura GA, Bloss JD. Prognostic factors 25. Tewari KS, Monk BJ. Beyond platinum al. The human papillomavirus type 18 E6
for response to cisplatin-based chemother- for metastatic and recurrent carcinoma of oncoprotein induces vascular endothelial
apy in advanced cervical carcinoma: a Gy- the cervix. Onkologie 2009;32:552-4. growth factor 121 (VEGF121) transcrip-
necologic Oncology Group study. Gynecol 26. Mantel N. Evaluation of survival data tion from the promoter through a p53-
Oncol 2010;116:44-9. and two new rank order statistics arising independent mechanism. Exp Cell Res
17. Tewari KS, Monk BJ. Recent achieve- in its consideration. Cancer Chemother 2007;313:3239-50.
ments and future developments in advanced Rep 1966;50:163-70. 37. Bodily JM, Mehta KP, Laimins LA. Hu-
and recurrent cervical cancer: trials of the 27. Cox DR. Regression models and life- man papillomavirus E7 enhances hypoxia-
Gynecologic Oncology Group. Semin Oncol tables. J R Stat Soc [B] 1972;34:187-220. inducible factor 1-mediated transcription
2009;36:170-80. 28. Wieand S, Schroeder G, O’Fallon JR. by inhibiting binding of histone deacety-
18. Leung DW, Cachianes G, Kuang WJ, Stopping when the experimental regimen lases. Cancer Res 2011;71:1187-95.
Goeddel DV, Ferrara N. Vascular endothe- does not appear to help. Stat Med 38. Monk BJ, Mas Lopez L, Zarba JJ, et al.
lial growth factor is a secreted angiogenic 1994;13:1453-8. Phase II, open-label study of pazopanib or
mitogen. Science 1989;246:1306-9. 29. Lan KK, DeMets DL. Discrete sequential lapatinib monotherapy compared with
19. Monk BJ, Sill MW, Burger RA, Gray boundaries for clinical trials. Biometrika pazopanib plus lapatinib combination
HJ, Buekers TE, Roman LD. Phase II trial 1983;70:659-63. therapy in patients with advanced and re-
of bevacizumab in the treatment of per- 30. Diggle P, Liang KY, Zeger SL. Analysis current cervical cancer. J Clin Oncol
sistent or recurrent squamous cell carci- of longitudinal data. Oxford, United King- 2010;28:3562-9.
noma of the cervix: a Gynecologic Oncol- dom: Clarendon Press, 1993. Copyright © 2014 Massachusetts Medical Society.
my nejm in the journal online
Individual subscribers can store articles and searches using a feature
on the Journal’s website (NEJM.org) called “My NEJM.”
Each article and search result links to this feature. Users can create
personal folders and move articles into them for convenient retrieval later.
n engl j med 370;8 nejm.org february 20, 2014 743
The New England Journal of Medicine
Downloaded from nejm.org at DIGNITY HEALTH on February 19, 2014. For personal use only. No other uses without permission.
Copyright © 2014 Massachusetts Medical Society. All rights reserved.You can also read
