A VELOUR post hoc subset analysis: prognostic groups and treatment outcomes
in patients with metastatic colorectal cancer treated with aflibercept and FOLFIRI
Ian Chau1, Florence Joulain2, Sheikh Usman Iqbal3, John Bridgewater4
1Department
of Medicine, Royal Marsden Hospital, London & Surrey, United Kingdom
2Sanofi,
Research and Development, Chilly Mazarin, France
3Sanofi,
Cambridge, Massachusetts, United States
4Department
of Oncology, University College London Hospitals, London, United
Kingdom
Corresponding Author:
Ian Chau, MD
Department of Medicine
Royal Marsden Hospital
Sutton, Surrey, UK, SM2 5PT
Phone: +44-208-915-6196
Fax: +44-208-661-3890
Email: Ian.Chau@rmh.nhs.uk
Affiliations and Contact Information for coauthors
Florence Joulain, MSc
Sanofi
Research and Development
1
1 av. Pierre Brossolette
Chilly Mazarin, France 91385
Phone: +33-1-53-77-88-29
E-mail: Florence.joulain@sanofi.com
Sheikh Usman Iqbal, MD, MPH, MBA
Sanofi
500 Kendall Street
Cambridge, MA, USA 02142
Phone: 617-768-1913
E-mail: Usman.Iqbal@sanofi.com
John Bridgewater, MD, PhD
Department of Oncology
University College London Hospitals
London, UK
Phone: 020-3447-9093
E-mail: j.bridgewater@ucl.ac.uk
2
ABSTRACT
Background:
The
VELOUR
study
demonstrated
a
survival
benefit
for
FOLFIRI+aflibercept versus FOLFIRI+placebo in metastatic colorectal cancer (mCRC)
patients who progressed on oxaliplatin-based chemotherapy. Continued divergence of
overall survival (OS) curves in the intension to treat (ITT) population, with the survival
advantage persisting beyond median survival time, suggested subpopulations might
have different magnitudes of survival gain. Additionally, 10% of patients within VELOUR
had recurrence during or within 6 months of completing oxaliplatin-based adjuvant
therapy (adjuvant fast relapsers) - previously identified as having poorer survival
outcomes.
Methods: To determine which patients received the greatest benefit from FOLFIRIaflibercept, a post hoc multivariate analysis of the VELOUR ITT population was
conducted. Prognostic factors identified were applied to the ITT population, excluding
adjuvant fast relapsers, to derive OS prognostic profiles.
Results: The better efficacy subgroup was identified as patients within VELOUR
exclusive of adjuvant fast relapsers and had performance status (PS) 0 with any
number of metastatic site or PS 1 with <2 metastatic site. A significant improvement in
efficacy outcome was observed with aflibercept in the better efficacy subgroup. Median
OS for FOLFIRI-aflibercept and FOLFIRI-placebo:16.2 and 13.1 months (adjusted
Hazard Ratio [HR]=0.73; 95% confidence interval [CI]: 0.61–0.86); median progression
free survival (PFS): 7.2 and 4.8 months (adjusted HR=0.68; 95% CI: 0.57-0.80); and
objective response rate (ORR): 24% versus 11% respectively. Poorer efficacy subgroup
3
comprised of adjuvant fast relapsers or patients with PS2 or PS1 with ≥2 metastatic
sites. In poorer efficacy subgroup, no benefit was seen with aflibercept. Median OS for
FOLFIRI-aflibercept and FOLFIRI-placebo: 10.4 and 9.6 months (adjusted HR=0.97;
95% CI: 0.78-1.21) respectively with no improvement in PFS or ORR.
Conclusion: This analysis suggests that within VELOUR, patients in the better efficacy
subgroup may derive enhanced benefit from treatment with FOLFIRI-aflibercept. These
prognostic criteria may guide practitioners toward optimal use of targeted biologicals in
appropriate second-line mCRC patients.
Key Words
Aflibercept, Metastatic colorectal cancer, Chemotherapy, Anti-VEGF agents, Targeted
therapy
4
Background
Colorectal cancer (CRC) accounts for more than 1.2 million new cancer cases and over
600,000 deaths [1]. Among these cases, approximately 25% of patients present with
metastatic CRC (mCRC) at diagnosis and up to half of all patients with CRC will
develop metastases during the course of their disease [2]. Metastatic unresectable CRC
is generally incurable and has a 5-year survival rate of only 10-15% [3,4]. In most
patients the disease will progress on initial treatment, and up to half of all patients with
mCRC will receive second- and third-line treatments [2,5]. Better second-line treatments
for mCRC are currently needed in this crucial area of high unmet medical need.
A number of treatment options exist for patients whose disease progresses after initial
therapy [2,6,7], and current real-world treatment patterns show that targeted agents plus
standard second-line chemotherapy are commonly utilised [5]. Aflibercept (known as
ziv-aflibercept in the United States) is a novel fusion protein containing vascular
endothelial growth factor (VEGF)–binding portions from the extracellular domains of
human VEGF receptors 1 and 2, fused to the Fc portion of human immunoglobulin (Ig)
G1. Aflibercept blocks the activity of VEGFA, VEGFB, and placental growth factor
(PIGF) by acting as a high-affinity ligand trap, preventing these ligands from binding to
their endogenous receptors [8]. Aflibercept is a new option in the second-line treatment
of mCRC and compared to placebo, it has demonstrated statistically significant
improvements in overall survival (OS) (median 13.50 versus 12.06 months; hazard ratio
[HR] = 0.817, p = 0.0032), progression-free survival (PFS) (6.90 versus 4.67 months;
HR = 0.758, p < 0.0001), and overall response rate (ORR) [19.8% (95% confidence
5
interval [CI]: 16.4-23.2) versus 11.1% (95% CI: 8.5-13.8); p < 0.0001] when used in
combination with FOLFIRI (5-fluorouracil-leuocovorin-irinotecan) in patients with mCRC
that is resistant to or has progressed after an oxaliplatin-containing regimen [9]. The
pre-specified subgroup analyses in the VELOUR trial showed a consistent benefit of
aflibercept across all major subgroups, including patients who had previously received
bevacizumab treatment [10]. Continued divergence of OS curves in the intention to treat
(ITT) population, with the survival advantage persisting beyond median survival time,
suggested subpopulations might have different magnitudes of survival gain.
The identification of patients who are likely to derive more benefit from targeted antiangiogenic therapies has become increasingly crucial. Despite more than a decade of
intense research effort, no tissue-based predictive biomarkers have yet been
established for anti-angiogenic therapy. An optimised treatment approach involving
careful patient selection based on identified prognostic factors and known tumor
characteristics may enable a more tailored approach to individual treatment. A growing
body of evidence indicates that patient characteristics, such as age, performance
status, time to recurrence following adjuvant therapy, extent of metastases, and use of
prior therapies, can all have an impact on treatment efficacy in mCRC [2]. With these
advances in the knowledge of prognostic factors in mCRC, there exists an opportunity
in the real-world setting to target patient populations based on clinically relevant
phenotypes and prognostic variables. The objective of this post hoc subgroup analysis
described herein was to identify a “better efficacy” patient population within VELOUR
study that may have derived greater clinical benefit from treatment with FOLFIRI plus
aflibercept.
6
Methods
This was a post hoc analysis of data from the VELOUR clinical trial which assessed
1,226 patients with histologically or cytologically proven colorectal adenocarcinoma and
who had previously received and progressed following oxaliplatin-based chemotherapy.
[9]. A pre-specified multivariate analysis was conducted on the ITT population in the
VELOUR trial to identify prognostic factors associated with improved OS.
Outcome measures
The primary end point of this analysis was OS in the better and poorer efficacy
subgroup of patients identified in the multivariate analysis. OS was defined as the time
interval from randomisation to death from any cause. PFS was defined as the time
interval from randomisation to tumor progression or death from any cause, whichever
comes first. Response was assessed according to Response Evaluation Criteria In
Solid Tumors (RECIST) version 1.0 [11] based on independent radiology review,
blinded to patient treatment.
Statistical analyses
Efficacy analyses were conducted in the randomised population according to the arm to
which patients were assigned at randomisation; response rate was assessed in the
subset of these patients who had measurable disease at baseline. Safety was analysed
using descriptive methods, in the treated population and subgroups, according to the
treatment received (patients who received ≥1 dose of aflibercept were analyzed in the
aflibercept arm, regardless of treatment assignment). Time-to-event parameters were
7
estimated using Kaplan-Meier analysis [12]. The hazard ratio (HR) and 95% confidence
interval (CI) estimates were provided using a Cox proportional hazards model [13].
Covariates used were based on pre-specified subgroups, including: Eastern
Cooperative Oncology Group [ECOG] performance status (PS), prior bevacizumab
status, age, prior hypertension, and number of metastatic organs involved. Formal
statistical interaction testing was performed between treatment and different prognostic
factors at the 2-sided 10% level.
8
Results
Identification of the better efficacy subgroup
The multivariate analysis identified PS and number of metastatic sites as key prognostic
factors influencing OS in the VELOUR trial (Table 1). Based on the standardised
literature review, the subgroup of patients who relapsed at or within 6 months of
adjuvant oxaliplatin-based chemotherapy (adjuvant fast relapsers; n= 124 patients;
~10% of ITT population) were excluded in view of their poor prognosis [14]. At the time
of regulatory approval by European Medicine Agency, a survival analysis excluding
adjuvant fast relapsers was specifically requested by the Committee for Medicinal
Products for Human Use and these results are included in the Summary of Product
Characteristics.
Prognostic profiles associated with OS, i.e. PS and number of metastases (identified
from the pre-specified multivariate analysis), were applied to the ITT population after
exclusion of adjuvant fast relapsers. To test the robustness of this prognostic model,
the survival of adjuvant fast relapsers and the non-adjuvant fast relapsers within
VELOUR was examined according to treatment arm. In non-adjuvant fast relapsers, the
HR was 0.78 (95%CI: 0.68-0.90) in favour of aflibercept whereas for adjuvant fast
relapsers, HR was 1.12 (95%CI: 0.72-1.74) with no survival benefit seen with
aflibercept. A qualitative interaction (although not statistically significant) was noted
between treatment and adjuvant fast relapsers likely due to the relative small number
(10%) of adjuvant fast relapsers within the VELOUR ITT population.
9
In the ITT excluding adjuvant fast relapsers population, the interaction between the
number of organs with metastases and treatment was significant at the 10% level in the
PS1 subgroup (p = 0.0725), whereas no interaction was found in the PS0 subgroup (p =
0.5151), indicating that the treatment effect is different according to the number of
metastatic sites in patients with PS1 but not in patients with PS0. Based on the strength
of association of the covariates, patients with a performance status of PS0 or PS1 with
<2 metastatic site were defined as better efficacy subgroup.
Figure 1 shows the
construction of the prognostic profile for the better efficacy subgroup. Therefore the
final better efficacy subgroup was comprised of patients within VELOUR study exclusive
of adjuvant fast relapsers and had either PS0 with any number of metastatic sites or
PS1 with <2 metastatic site. The interaction between treatment and the better efficacy
subgroup was statistically significant (p=0.0147) indicating that there was a differential
OS effect of aflibercept in this better efficacy subgroup compared to placebo. The
poorer efficacy subgroup comprised of patients within VELOUR study who were
adjuvant fast relapsers or PS 1 with ≥2 metastatic sites or PS2 with any number of
metastatic sites.
The better efficacy subgroup consisted of 406 patients (66%) from the FOLFIRI-placebo
control arm and 404 patients (66%) from the FOLFIRI-aflibercept investigational arm of
the VELOUR trial. Table 2 shows baseline characteristics, which were similar in both
study arms. Within the better efficacy subgroup, 67% of patients received no prior
treatment with bevacizumab.
10
Efficacy
Figure 2A shows OS for the better efficacy subgroup by treatment arms. In the better
efficacy subgroup, 293/406 patients (72.2%) receiving FOLFIRI-placebo died during the
observation period compared with 244/404 patients (60.4%) receiving FOLFIRIaflibercept. There was a significant improvement in OS with aflibercept in the better
efficacy subgroup (adjusted HR: 0.73; 95% CI: 0.61-0.86). The median OS was 13.1
months (95% CI: 11.7-14.2) for FOLFIRI-placebo-treated patients and 16.2 months
(95% CI: 14.5-18.1) for those treated with FOLFIRI-aflibercept with an absolute
difference of 3.1 months in median OS.
Over time, the magnitude of survival
differences between the 2 treatment groups continued to increase in the better efficacy
subgroup with the absolute OS rate differences increased from 5% at 6 months to 15%
at 30 months (Figure 2A). Table 3 shows a summary of efficacy results for better and
poorer efficacy subgroups by treatment arms.
Figure 2B shows the OS for the poorer efficacy subgroup by treatment arms.
By
contrast, in the poorer efficacy subgroup, no benefit was seen with aflibercept. 167/208
(80.3%) patients receiving FOLFORI-placebo died during the observation period versus
159/208 (76.4%) receiving FOLFIRI-aflibercept. The adjusted HR for FOLFIRIaflibercept versus FOLFIRI-placebo was 0.97 (95%CI: 0.78-1.21). The median OS was
10.4 months (95% CI: 9.5-12.1) for FOLFIRI-placebo and 9.6 months (95% CI: 8.6-11.5)
for FOLFIRI-aflibercept. There were no notable absolute survival differences between
the two treatment arms in the poorer efficacy subgroup (Figure 2B and Table 3). Figure
3 shows a forest plot of overall survival for the better and poorer efficacy subgroups.
11
Survival was similar in the better efficacy subgroup irrespective of prior bevacizumab
use.
In addition, for the better efficacy subgroup, a significant improvement in PFS was also
observed for FOLFIRI-aflibercept. The adjusted HR for FOLFIRI-aflibercept versus
FOLFIRI-placebo for PFS was 0.68 (95% CI: 0.57-0.80). Figure 4 shows the PFS for the
better efficacy subgroup by treatment arms. The median PFS for FOLFIRI-placebo was
4.8 months (95% CI: 4.2-5.4) compared with 7.2 months for FOLFIRI-aflibercept (95%
CI: 6.8-8.2), an absolute difference of 2.4 months in median PFS.
difference in 6-month PFS rates was 25%.
The absolute
Similarly, for the better efficacy subgroup,
the objective response rate (ORR) results followed a consistent pattern: 23.7% (95% CI:
19.3-28.2) for FOLFIRI-aflibercept and 11% for FOLFIRI-placebo (95% CI: 7.8-14.3).
On the contrary, no improvements in PFS or ORR were observed with aflibercept in the
poorer efficacy subgroup.
Dose intensities were similar between better and poorer efficacy groups and therefore
could not account for the differential survival effect of aflibercept (Supplementary Table
1). However, as the better efficacy group had a longer PFS, the mean number of cycles
of aflibercept was more for the better efficacy subgroup.
Safety and tolerability
Supplementary Table 2 shows adverse reactions and abnormalities in laboratory values
(all grades) reported at a higher incidence (≥2%) in patients treated with aflibercept plus
FOLFIRI compared with placebo plus FOLFIRI in the overall safety population as well
12
as the better and poorer efficacy subgroups. No notable differences were seen with all
grades and grade 3 /4 AEs for the better and poorer efficacy subgroups compared with
the overall safety population as well as between the two subgroups.
In the better efficacy subgroup, the most commonly encountered AEs were
gastrointestinal in the FOLFIRI-aflibercept group as was also seen in the entire study
population. Diarrhoea (all grades) occurred in approximately 70% of patients in the
FOLFIRI-aflibercept group and in approximately 50% of those in the FOLFIRI-placebo
group. Leucopenia (all grades) occurred in 78.5% of patients in the FOLFIRI-aflibercept
group and in approximately 73% of those in the FOLFIRI-placebo group. Stomatitis and
hypertension were more common with FOLFIRI-aflibercept than with FOLFIRI-placebo.
Severe AEs (≥ grade 3) reported in the FOLFIRI-aflibercept arm included diarrhoea,
stomatitis, fatigue, hypertension, and neutropenia. However, the majority of these
severe treatment-associated AEs were grade 3 and included known VEGF-inhibitor
class effects, such as hypertension, proteinuria, and haemorrhage, and known
chemotherapy-associated
AEs,
such
as
diarrhoea,
stomatitis,
infection,
and
neutropenia.
As in the overall safety population and the better efficacy subgroup, in the poorer
efficacy subgroup leucopenia was the most commonly encountered AE with FOLFIRIaflibercept. Diarrhoea occurred in approximately 65% in both groups. Severe AEs (≥
grade 3) reported in the FOLFIRI-aflibercept arm included diarrhoea, stomatitis, fatigue,
hypertension, and neutropenia. The incidence of grade ≥3 treatment-emergent AEs
provided were consistent between the 2 subgroups, with slightly more grade ≥3
13
vomiting, decreased appetite, and weight loss in the poorer responders subgroup and
more grade ≥3 proteinuria in the better efficacy subgroup.
14
Discussion
The VELOUR trial demonstrated a statistically significant benefit of adding aflibercept to
FOLFIRI in terms of OS, PFS, and ORR when compared with placebo plus FOLFIRI in
the ITT population. This post hoc survival analysis of the VELOUR study has identified a
better efficacy subgroup that may have an enhanced benefit from the addition of
aflibercept to FOLFIRI. Data from the post hoc better efficacy subgroup demonstrated
that patients who received FOLFIRI-aflibercept had a median OS of 16.2 months (95%
CI: 14.5-18.1) compared with 13.1 months (95% CI: 11.7-14.2) for FOLFIRI-placebo, a
difference of 3.1 months. The adjusted HR versus FOLFIRI-placebo was 0.73 (95%CI:
0.61-0.86] as compared to 0.82 (95%CI: 0.71-0.94) in the ITT population. In the better
efficacy subgroup, survival differences increased over time reaching 15% at 30 months.
The results for PFS and ORR also followed a consistent pattern of improvement in the
FOLFIRI-aflibercept regimen over the FOLFIRI-alone regimen.
The efficacy outcome benefits in the better efficacy subgroup were obtained without any
compromise to safety, where the AEs in the better efficacy subgroup closely mirrored
those in the overall safety population. The toxicity profile is in line with expectations for
the anti-VEGF plus chemotherapy class effect seen in other studies.
With regard to metastases, the primary pre-specified analysis in the ITT population
showed that patients with liver-only metastases had a better treatment efficacy. In our
post hoc analysis, the prognostic factor identified related to the number of metastases
(≤1 metastasis) and was not limited to liver-only metastasis or specific metastasis and
sufficed for any organ involvement. Data from current post-hoc analysis, however, did
15
provide further rationale to evaluate aflibercept in patients with liver only metastases
who require downsizing conversion chemotherapy.
Subgroup analyses in clinical trials are becoming increasingly important to demonstrate
the consistency of the treatment effect and may provide guidance to practising
physicians. Assessment of subgroups is usually confined to pre-specified analyses and
a priori identification at the beginning of a trial. However, given the wide heterogeneity
of treatment effect in relation to tumor biology and pathophysiology in oncology, welldesigned post hoc analyses in clinical trials aligned with biologic plausibility should be
considered an important means to explore treatment benefits in clinically relevant
subpopulations.
The findings of this study are aligned with other published data in terms of the
assessment of clinical determinants of survival in mCRC. Köhne et al. analyzed 3,825
patients with mCRC who received treatment with 5-fluorouracil in 19 prospective,
randomized, controlled trials, and identified 3 prognostic risk groups based on 4 clinical
parameters that included both PS and number of metastatic sites [15]. PS0 or PS1 was
found to be a positive predictor of overall survival, whereas more than 1 metastatic site
was a negative predictor of overall survival. Díaz et al. subsequently confirmed the
application of the Köhne model in a retrospective study of an mCRC patient population
treated with an irinotecan-based or oxaliplatin-based chemotherapy and concluded that
ECOG PS and the number of metastatic sites were 2 of the key prognostic factors for
predicting OS [16].
16
The limitations of this study were similar to other published post hoc analyses. The
findings were a result of a non a priori analysis, and hence could not rule out selection
bias and multiplicity aspects relating to a post hoc analytic exercise. Thus this post hoc
survival analysis was only hypothesis-generating and the findings need confirmation in
prospective trials.
In conclusion, the better efficacy subgroup, comprising of patients within VELOUR study
exclusive of adjuvant fast relapsers and had either PS0 with any number of metastatic
sites or PS1 with <2 metastatic site, might derive enhanced efficacy benefits (OS, PFS,
and ORR) with the combination of aflibercept and FOLFIRI compared with FOLFIRI
alone. The magnitude of OS benefit in the better efficacy subgroup appeared
numerically higher than that seen in the ITT population. The identification of this better
efficacy subgroup profile for second-line patients with mCRC may help to guide
practitioners toward targeted use of biologicals for maximal survival gains in clinically
relevant patient populations.
17
Abbreviations
AEs, adverse events; CI, confidence interval; CRC, colorectal cancer; ECOG, Eastern
Cooperative Oncology Group; FOLFIRI, 5-fluorouracil-leuocovorin-irinotecan; HR,
hazard ratio; ITT, intent-to-treat; IV, intravenous; mCRC, metastatic colorectal cancer;
ORR, overall response rate; OS, overall survival; PFS, progression-free survival; PS,
performance status; VEGF, vascular endothelial growth factor
Competing Interests
The manuscript was financially supported by Sanofi (Cambridge, MA) and Regeneron
Pharmaceuticals, Inc. (Tarrytown, NY). The Medical Affairs Departments at Sanofi and
Regeneron were allowed several courtesy reviews by the authors for scientific
accuracy, and provided feedback to the authors for their consideration. Sanofi and
Regeneron were not involved in the writing or editing of this manuscript, and were not
permitted to censor any content from the authors.
Authors’ Contributions
All authors meet the criteria for authorship as recommended by the International
Committee of Medical Journal Editors (ICMJE), were fully responsible for all content and
editorial decisions, retained full control over all content contained in this manuscript, and
were involved with all stages of manuscript development.
IC, FJ, SUI and JB were involved in study concept and design. FJ performed the
statistical analysis. IC, FJ, SUI and JB were involved in the drafting , intellectual input
and final approval of the manuscript.
18
Acknowledgements
Editorial and writing assistance in the development of this manuscript in the form of
drafting and revising content based on specific direction from the authors, collation of
author comments, editing, referencing, manuscript formatting, and creation of tables
and figures was provided by Connexion Healthcare (Newtown, PA).
IC would like to acknowledge National Health Service funding to the National Institute
for Health Research Biomedical Research Centre
19
REFERENCES
1. Ferlay J, Shin HR, Bray F, Forman D, Mathers C, Parkin DM. GLOBOCAN 2008,
Cancer Incidence and Mortality Worldwide: IARC CancerBase No. 10 [Internet].
Lyon, France: International Agency for Research on Cancer. 2010.
2. Schmoll HJ, Van CE, Stein A, Valentini V, Glimelius B, Haustermans K et al.:
ESMO Consensus Guidelines for management of patients with colon and rectal
cancer. a personalized approach to clinical decision making. Ann Oncol 2012,
23: 2479-2516.
3. Masi G, Vasile E, Loupakis F, Cupini S, Fornaro L, Baldi G et al.: Randomized
trial of two induction chemotherapy regimens in metastatic colorectal cancer: an
updated analysis. J Natl Cancer Inst 2011, 103: 21-30.
4. Sanoff HK, Sargent DJ, Campbell ME, Morton RF, Fuchs CS, Ramanathan RK et
al.: Five-year data and prognostic factor analysis of oxaliplatin and irinotecan
combinations for advanced colorectal cancer: N9741. J Clin Oncol 2008, 26:
5721-5727.
5. Hess GP, Wang PF, Quach D, Barber B, Zhao Z: Systemic Therapy for
Metastatic Colorectal Cancer: Patterns of Chemotherapy and Biologic Therapy
Use in US Medical Oncology Practice. J Oncol Pract 2010, 6: 301-307.
20
6. Benson AB, III, Bekaii-Saab T, Chan E, Chen YJ, Choti MA, Cooper HS et al.:
Rectal cancer. J Natl Compr Canc Netw 2012, 10: 1528-1564.
7. Benson AB, III, Bekaii-Saab T, Chan E, Chen YJ, Choti MA, Cooper HS et al.:
Metastatic colon cancer, version 3.2013: featured updates to the NCCN
Guidelines. J Natl Compr Canc Netw 2013, 11: 141-152.
8. Holash J, Davis S, Papadopoulos N, Croll SD, Ho L, Russell M et al.: VEGFTrap: a VEGF blocker with potent antitumor effects. Proc Natl Acad Sci U S A
2002, 99: 11393-11398.
9. Van CE, Tabernero J, Lakomy R, Prenen H, Prausova J, Macarulla T et al.:
Addition of aflibercept to fluorouracil, leucovorin, and irinotecan improves survival
in a phase III randomized trial in patients with metastatic colorectal cancer
previously treated with an oxaliplatin-based regimen. J Clin Oncol 2012, 30:
3499-3506.
10. Tabernero J, Van CE, Lakomy R, Prausova J, Ruff P, van Hazel GA et al.:
Aflibercept versus placebo in combination with fluorouracil, leucovorin and
irinotecan in the treatment of previously treated metastatic colorectal cancer:
prespecified subgroup analyses from the VELOUR trial. Eur J Cancer 2014, 50:
320-331.
21
11. Therasse P, Arbuck SG, Eisenhauer EA, Wanders J, Kaplan RS, Rubinstein L et
al.: New guidelines to evaluate the response to treatment in solid tumors.
European Organization for Research and Treatment of Cancer, National Cancer
Institute of the United States, National Cancer Institute of Canada. J Natl Cancer
Inst 2000, 92: 205-216.
12. Kaplan EL, Meier P: Non parametric estimation from incomplete observations.
Journal of American Statistical Association 1958, 53: 457-481.
13. Cox DR: Regression models and life tables. J R Stat Soc A 1972, 29: 187-220.
14. O'Connell MJ, Campbell ME, Goldberg RM, Grothey A, Seitz JF, Benedetti JK et
al.: Survival following recurrence in stage II and III colon cancer: findings from the
ACCENT data set. J Clin Oncol 2008, 26: 2336-2341.
15. Kohne CH, Cunningham D, Di CF, Glimelius B, Blijham G, Aranda E et al.:
Clinical determinants of survival in patients with 5-fluorouracil-based treatment
for metastatic colorectal cancer: results of a multivariate analysis of 3825
patients. Ann Oncol 2002, 13: 308-317.
16. Diaz R, Aparicio J, Girones R, Molina J, Palomar L, Segura A et al.: Analysis of
prognostic factors and applicability of Kohne's prognostic groups in patients with
metastatic colorectal cancer treated with first-line irinotecan or oxaliplatin-based
chemotherapy. Clin Colorectal Cancer 2005, 5: 197-202.
22
Table 1. Results from the pre-specified multivariate Cox model
Hazard ratio
Prognostic factors
(95.34% CI)
p-value
0.817 [0.713; 0.936]
0.0032
ECOG PS: 1 vs 0
1.683 [1.465; 1.935]
< 0.0001
ECOG PS: 2 vs 0
3.625 [2.395; 5.486]
<0 .0001
Prior bevacizumab: yes vs no
1.165 [1.003; 1.353]
0.0425
Age (years): ≥65 vs <65
1.179 [1.020; 1.363]
0.0233
Prior hypertension: yes vs no
0.785 [0.682; 0.904]
0.0007
1.446 [1.258; 1.663]
< 0.0001
Treatment: FOLFIRI-aflibercept versus
FOLFIRI-placebo
Number of metastatic organs involved
as per IRC: >1 vs (0-1)
ECOG, European Cooperative Oncology Group; FOLFIRI, 5-fluorouracil–leucovorin–
irinotecan; IRC, independent review committee; PS, performance status.
23
Table 2. Baseline characteristics of patients in the post hoc subgroup analysis
Better efficacy subgroup
ITT population
FOLFIRI-
FOLFIRI-
FOLFIRI-
FOLFIRI-
Placebo
Aflibercept
Placebo
Aflibercept
(n = 406)
(n = 404)
(n = 614)
(n = 612)
Female
174 (42.9)
153 (37.9)
261 (42.5)
247 (40.4)
Male
232 (57.1)
251 (62.1)
353 (57.5)
365 (59.6)
59.9 (25-86)
59.4 (21-82)
<65 y
254 (62.6)
277 (68.6)
376 (61.2)
407 (66.5)
≥65 but <75 y
133 (32.8)
108 (26.7)
199 (32.4)
172 (28.1)
19 (4.7)
19 (4.7)
39 (6.4)
33 (5.4)
Asian
29 (7.1)
23 (5.7)
51 (8.3)
35 (5.7)
Black
14 (3.4)
11 (2.7)
27 (4.4)
16 (2.6)
White
355 (87.4)
361 (89.4)
523 (85.2)
548 (89.5)
Other
8 (2)
9 (2.2)
13 (2.1)
13 (2.1)
Colon
187 (46.1)
188 (46.5)
302 (49.2)
289 (47.2)
Rectosigmoid
95 (23.4)
85 (21)
136 (22.1)
123 (20.1)
Rectum
123 (30.3)
129 (31.9)
174 (28.3)
197 (32.2)
1 (0.2)
2 (0.5)
2 (0.3)
3 (0.5)
Parameter
Gender, n (%)
Mean age, y (range)
60.2 (19-86) 59.5 (21-82)
Age group, n (%)
≥75 y
Race, n (%)
Cancer diagnosis
category, n (%)
Other
24
Number of organs with
metastasis, n (%)
0
2 (0.5)
2 (0.5)
6 (1.0)
2 (0.3)
1
237 (58.4)
230 (56.9)
271 (44.1)
256 (41.8)
>1
167 (41.1)
172 (42.6)
337 (54.9)
354 (57.8)
Liver
283 (69.7)
294 (72.8)
431 (70.2)
459 (75.0)
Lung
160 (39.4)
152 (37.6)
277 (45.1)
271 (44.3)
Lymph
88 (21.7)
100 (24.8)
181 (29.5)
173 (28.3)
134 (33)
141 (34.9)
146 (23.8)
153 (25.0)
0
314 (77.3)
317 (78.5)
350 (57.0)
349 (57.0)
1
92 (22.7)
87 (21.5)
250 (40.7)
250 (40.8)
2
0
0
14 (2.3)
13 (2.1)
Yes
133 (32.8)
132 (32.7)
187 (30.5)
186 (30.4)
No
273 (67.2)
272 (67.3)
427 (69.5)
426 (69.6)
Metastatic sites, n (%)
Liver metastasis only, n
(%)
ECOG performance
status, n (%)
Prior bevacizumab, n (%)
ECOG, European Cooperative Oncology Group; FOLFIRI, 5-fluorouracil–leucovorin–
irinotecan; ITT, intent to treat; PS, performance status.
25
Table 3 Summary of efficacy results for better and poorer efficacy subgroups by
treatment arms.
Better efficacy subgroup
Poor efficacy subgroup
FOLFIRIplacebo
FOLFIRIaflibercept
FOLFIRIplacebo
FOLFIRIaflibercept
6 months
83%
88%
71%
70%
12 months
54%
64%
43%
42%
18 months
34%
46%
24%
24%
24 months
22%
35%
12%
15%
30 months
12%
27%
10%
13%
Overall survival rates
Progression free survival rates
6 months
38%
63%
12 months
15%
20%
18 months
5%
6%
Objective response rates
11%
24%
26
Supplementary Table 1 Dose intensities for better and poorer efficacy subgroups
by treatment arms
Better efficacy subgroup
Number of cycles received by patient
Sum
Mean (SD)
Median
Min : Max
1
2
3
4
5
6
7
8
9
10
11-15
16-20
21-25
>25
Duration of exposure to aflibercept/placebo
(weeks)
Number
Mean (SD)
Median
Min : Max
Total cumulative dose received (mg/kg)
Number
Mean (SD)
Median
Min : Max
Placebo/Folfiri
(N=401)
Aflibercept/Folfiri
(N=407)
4103
10.2 (7.8)
9.0
1 : 67
4045
9.9 (7.3)
8.0
1 : 35
10 (2.5%)
17 (4.2%)
61 (15.2%)
19 (4.7%)
18 (4.5%)
29 (7.2%)
19 (4.7%)
22 (5.5%)
30 (7.5%)
15 (3.7%)
84 (20.9%)
40 (10.0%)
17 (4.2%)
20 (5.0%)
22 (5.4%)
28 (6.9%)
42 (10.3%)
25 (6.1%)
29 (7.1%)
20 (4.9%)
18 (4.4%)
22 (5.4%)
21 (5.2%)
20 (4.9%)
71 (17.4%)
48 (11.8%)
26 (6.4%)
15 (3.7%)
401
22.9 (17.3)
19.0
2 : 135
407
23.4 (17.0)
19.0
2 : 85
401
40.73 (31.31)
36.00
0.6 : 266.4
407
38.36 (28.69)
32.00
3.8 : 140.0
27
Better efficacy subgroup
Actual dose intensity (mg/kg/week)
Number
Mean (SD)
Median
Min : Max
Relative dose intensity
Number
Mean (SD)
Median
Min : Max
Poorer efficacy subgroup
Number of cycles received by patient
Sum
Mean (SD)
Median
Min : Max
1
2
3
4
5
6
7
8
9
10
11-15
16-20
21-25
>25
Placebo/Folfiri
(N=401)
Aflibercept/Folfiri
(N=407)
401
1.78 (0.25)
1.84
0.3 : 2.1
407
1.53 (0.44)
1.65
0.1 : 2.1
401
0.89 (0.13)
0.92
0.2 : 1.0
407
0.76 (0.22)
0.82
0.1 : 1.1
Placebo/Folfiri
(N=204)
Aflibercept/Folfiri
(N=204)
1932
9.5 (8.2)
7.0
1 : 45
1587
7.8 (6.7)
5.0
1 : 31
14 (6.9%)
15 (7.4%)
24 (11.8%)
12 (5.9%)
14 (6.9%)
16 (7.8%)
10 (4.9%)
12 (5.9%)
15 (7.4%)
6 (2.9%)
28 (13.7%)
17 (8.3%)
11 (5.4%)
10 (4.9%)
21 (10.3%)
24 (11.8%)
28 (13.7%)
20 (9.8%)
14 (6.9%)
9 (4.4%)
10 (4.9%)
7 (3.4%)
8 (3.9%)
8 (3.9%)
23 (11.3%)
20 (9.8%)
8 (3.9%)
4 (2.0%)
28
Placebo/Folfiri
(N=204)
Aflibercept/Folfiri
(N=204)
204
21.0 (18.0)
16.0
2 : 97
204
18.3 (15.5)
12.0
2 : 74
204
37.46 (32.28)
28.00
3.9 : 181.0
204
30.36 (25.68)
20.00
3.9 : 108.0
Actual dose intensity (mg/kg/week)
Number
Mean (SD)
Median
Min : Max
204
1.78 (0.24)
1.84
0.8 : 2.1
204
1.61 (0.43)
1.72
0.1 : 2.1
Relative dose intensity
Number
Mean (SD)
Median
Min : Max
204
0.89 (0.12)
0.92
0.4 : 1.1
204
0.81 (0.21)
0.86
0.1 : 1.0
Poorer efficacy subgroup
Duration of exposure to aflibercept/placebo
(weeks)
Number
Mean (SD)
Median
Min : Max
Total cumulative dose received (mg/kg)
Number
Mean (SD)
Median
Min : Max
29
Supplementary Table 2 Adverse reactions and abnormalities in laboratory values (all grades) reported at a higher
incidence (≥2%) in patients treated with aflibercept plus FOLFIRI compared with placebo plus FOLFIRI in the
overall safety population as well as the better and poorer efficacy subgroups.
Overall safety population
Better efficacy subgroup
Poorer efficacy subgroup
FOLFIRI-
FOLFIRI-
FOLFIRI-
FOLFIRI-
FOLFIRI-
FOLFIRI-
Primary system
placebo
aflibercept
placebo
aflibercept
placebo
aflibercept
organ class,
(n = 605)
(n = 611)
(n = 401)
(n = 407)
(n = 204)
(n = 204)
preferred term,
n (%)
All
Grades
All
Grades
All
Grades
All
Grades
All
Grades
All
Grades
grades
≥3
grades
≥3
grades
≥3
grades
≥3
grades
≥3
grades
≥3
37
5
56
5
25
3
36
3
12
2
20
2
(6.1)
(0.8)
(9.2)
(0.8)
(6.2)
(0.7)
(8.8)
(0.7)
(5.9)
(1.0)
(9.8)
(1.0)
Infections and infestations
Urinary tract
infection
15
Nasopharyngitis
28
0
(2.5)
12
0
(4.6)
23
-
(3.0)
3
-
(5.7)
5
-
(1.5)
(2.5)
Blood and lymphatic system disorders
432
73
472
94
291
51
317
55
141
22
155
39
(72.4)
(12.2)
(78.3)
(15.6)
(72.9)
(12.8)
(78.5)
(13.6)
(71.2)
(11.1)
(77.9)
(19.6)
336
176
409
221
227
119
276
142
109
57
133
79
Leukopeniaa
Neutropeniaa
30
(56.3)
(29.5)
(67.8)
(36.7)
(56.9)
(29.8)
(68.3)
(35.1)
(55.1)
(28.8)
(66.8)
(39.7)
202
10
286
20
147
8
195
14
55
2
91
6
(33.8)
(1.7)
(47.4)
(3.3)
(36.8)
(2.0)
(48.3)
(3.5)
(27.8)
(1.0)
(45.7)
(3.0)
10
10
26
26
6
6
18
18
4
4
8
8
(1.7)
(1.7)
(4.3)
(4.3)
(1.5)
(1.5)
(4.4)
(4.4)
(2.0)
(2.0)
(3.9)
(3.9)
144
11
195
21
92
6
129
10
52
5
66
11
(23.8)
(1.8)
(31.9)
(3.4)
(22.9)
(1.5)
(31.7)
(2.5)
(25.5)
(2.5)
(32.4)
(5.4)
18
8
55
26
7
4
31
16
11
4
24
10
(3.0)
(1.3)
(9.0)
(4.3)
(1.7)
(1.0)
(7.6)
(3.9)
(5.4)
(2.0)
(11.8)
(4.9)
53
2
136
10
36
1
97
8
17
1
39
2
(8.8)
(0.3)
(22.3)
(1.6)
(9.0)
(0.2)
(23.8)
(2.0)
(8.3)
(0.5)
(19.1)
(1.0)
65
9
252
117
43
5
178
80
22
4
74
37
(10.7)
(1.5)
(41.2)
(19.1)
(10.7)
(1.2)
(43.7)
(19.7)
(10.8)
(2.0)
(36.3)
(18.1)
Thrombocytopeniaa
Febrile neutropenia
Metabolic and nutritional disorders
Decreased appetite
Dehydration
Nervous system disorders
Headache
Vascular disorders
Hypertension
Respiratory, thoracic, and mediastinal disorders
31
45
169
1
32
(27.7)
(0.2)
(8.0)
155
3
11
125
1
13
(30.7)
(0.2)
(6.4)
110
3
9
44
–
Epistaxis
(7.4)
20
Dysphonia
0
(3.3)
–
(25.4)
(0.5)
(2.7)
–
(21.6)
45
(27.0)
(0.7)
(4.4)
–
(22.1)
52
5
72
5
28
1
45
3
24
4
27
2
(8.6)
(0.8)
(11.8)
(0.8)
(7.0)
(0.2)
(11.1)
(0.7)
(11.8)
(2.0)
(13.2)
(1.0)
46
1
9
13
1
(0.5)
Dyspnea
19
Oropharyngeal pain
(3.1)
(7.5)
11
38
Rhinorrhea
(1.8)
33
-
(0.2)
-
(2.2)
(8.1)
(4.9)
(6.4)
7
24
4
14
(6.2)
10
-
(1.7)
(5.9)
(2.0)
(6.9)
Gastrointestinal disorders
342
47
423
118
212
30
290
76
130
17
133
42
(56.5)
(7.8)
(69.2)
(19.3)
(52.9)
(7.5)
(71.3)
(18.7)
(63.7)
(8.3)
(65.2)
(20.6)
199
28
306
78
134
17
209
53
65
11
97
25
(32.9)
(4.6)
(50.1)
(12.8)
(33.4)
(4.2)
(51.4)
(13.0)
(31.9)
(5.4)
(47.5)
(12.3)
143
14
164
27
95
9
113
13
48
5
51
14
(23.6)
(2.3)
(26.8)
(4.4)
(23.7)
(2.2)
(27.8)
(3.2)
(23.5)
(2.5)
(25.0)
(6.9)
Diarrhea
Stomatitis
Abdominal pain
32
Abdominal pain
upper
48
6
66
7
33
5
47
2
15
1
19
5
(7.9)
(1.0)
(10.8)
(1.1)
(8.2)
(1.2)
(11.5)
(0.5)
(7.4)
(0.5)
(9.3)
(2.5)
13
Hemorrhoids
35
-
(2.1)
8
-
(5.7)
27
-
(2.0)
5
-
(6.6)
8
-
(2.5)
3.9
15
3
32
4
8
2
20
2
7
1
12
2
(2.5)
(0.5)
(5.2)
(0.7)
(2.0)
(0.5)
(4.9)
(0.5)
(3.4)
(0.5)
(5.9)
(1.0)
11
2
32
2
9
2
23
2
2
Rectal hemorrhage
Proctalgia
9
-
(1.8)
(0.3)
14
Aphthous stomatitis
(5.2)
(0.3)
(2.2)
30
4
13
(4.9)
(0.7)
(3.2)
(2.3)
5
Toothache
19
3
-
(3.1)
(0.5)
(1.0)
(4.4)
21
3
1
9
1
(5.2)
(0.7)
(0.5)
(4.4)
(0.5)
-
15
-
(0.7)
-
(5.7)
-
(0.8)
(0.5)
2
-
(3.7)
4
-
(1.0)
(2.0)
Skin and subcutaneous tissue disorders
Palmar-plantar
26
3
67
17
17
3
51
14
9
erythrodysesthesia
syndrome
16
3
(7.8)
(1.5)
11
-
(4.3)
(0.5)
(11.0)
(2.8)
(4.2)
(0.7)
(12.5)
(3.4)
(4.4)
17
-
50
-
11
-
39
-
6
Skin
-
hyperpigmentation
33
(2.8)
(8.2)
(2.7)
(9.6)
(2.9)
(5.4)
123
13
(60.3)
(6.4)
Renal and urinary disorders
246
7
380
48
165
7
257
35
81
(40.7)
(1.2)
(62.2)
(7.9)
(41.1)
(1.7)
(63.1)
(8.6)
(39.7)
108
3
136
66
2
90
Proteinuriab
Serum creatinine
-
increaseda
(18.1)
(0.5)
(22.6)
42
1
46
(21.2)
(0.5)
(23.4)
(16.6)
(0.5)
(22.3)
-
General disorders and administration site conditions
236
47
292
77
153
23
201
52
83
24
91
25
(39.0)
(7.8)
(47.8)
(12.6)
(38.2)
(5.7)
(49.4)
(12.8)
(40.7)
(11.8)
(44.6)
(12.3)
80
18
112
31
45
8
69
21
35
10
43
10
(13.2)
(3.0)
(18.3)
(5.1)
(11.2)
(2.0)
(17.0)
(5.2)
(17.2)
(4.9)
(21.1)
(4.9)
296
10
339
18
203
8
230
13
93
2
109
5
(50.2)
(1.7)
(57.5)
(3.1)
(51.8)
(2.0)
(57.8)
(3.3)
(47.0)
(1.0)
(56.8)
(2.6)
221
13
284
16
149
12
198
13
72
1
86
3
(37.1)
(2.2)
(47.3)
(2.7)
(37.5)
(3.0)
(49.1)
(3.2)
(36.4)
(0.5)
(43.7)
(1.5)
87
5
195
16
54
1
119
8
33
4
76
8
Fatigue
Asthenia
Investigations
AST increaseda
ALT
increaseda
Weight decrease
34
(14.4)
(0.8)
(31.9)
(2.6)
(13.5)
(0.2)
(29.2)
(2.0)
(16.2)
(2.0)
(37.3)
(3.9)
Abbreviations: FOLFIRI, 5-fluorouracil–leucovorin–irinotecan.
a
Based on laboratory values.
b
Compilation of clinical and laboratory data.
35
FIGURE LEGENDS
Figure 1 Construction of the prognostic profile for the better efficacy subgroup
ECOG, European Cooperative Oncology Group; OS, overall survival.
Figure 2A Overall survival for the better efficacy subgroup by treatment arms
Figure 2B Overall survival for the poorer efficacy subgroup by treatment arms
Figure 3 Forest plot of overall survival for the better and poorer efficacy
subgroups
CI, confidence interval
Adjusted HR: Hazard ratio adjusted on baseline value of performance status, prior
bevacizumab, age, hypertension and number of metastatic organs.
Figure 4 Progression free survival for the better efficacy subgroup by treatment
arms
36