E U RO P E A N U R O L O GY O N C O L O GY 2 ( 2 019 ) 7 0 8 – 715

available at www.sciencedirect.com
journal homepage: euoncology.europeanurology.com

EUO Priority Article

First-line Treatment of Metastatic Renal Cell Carcinoma:

A Systematic Review and Network Meta-analysis

Andrew W. Hahn a,y, Zachary Klaassen b,c,y, Neeraj Agarwal a,y, Benjamin Haaland a,
John Esther a, Xiang Y. Ye d, Xuechen Wang a, Sumanta K. Pal e,*, Christopher J.D. Wallis f,*
a
Division of Oncology, Department of Internal Medicine, Huntsman Cancer Institute, University of Utah, Salt Lake City, UT, USA; b Division of Urology, Medical
College of Georgia at Augusta University, Augusta, GA, USA; c Georgia Cancer Center, Augusta, GA, USA; d
MiCare Research Centre, Mount Sinai Hospital,
e
Toronto, Ontario, Canada; Department of Medical Oncology and Therapeutics Research, City of Hope Comprehensive Cancer Center, Duarte, CA, USA;
f
Division of Urology, Department of Surgery, University of Toronto, Toronto, Ontario, Canada

Article info Abstract

Article history: Context: No head-to-head clinical trials compare contemporary first-line therapies for
Received 7 July 2019Accepted metastatic renal cell carcinoma (mRCC). A network meta-analysis provides an approach
September 9, 2019 for quantitative analysis.
Objective: To indirectly compare the efficacy and safety of first-line treatments for
Associate Editor: Laurence Al- mRCC in the intention-to-treat (ITT) population and by clinical risk group.
biges Evidence acquisition: An updated systematic review from database inception to Febru-
ary 17, 2019 identified all parallel-group randomized controlled trials assessing first-line
Keywords: therapy for mRCC. “Clinically relevant” studies were selected for a network meta-analysis.
Metastatic renal cell carcinoma Progression-free survival (PFS) was the primary outcome. Overall survival (OS), overall
Network meta-analysis response rate (ORR), and grade 3 and 4 adverse events (AEs) were secondary outcomes.
International Metastatic Renal Evidence synthesis: We identified 12 relevant trials: 12 reported outcomes for PFS, nine
Cell Carcinoma Database for OS, 10 for ORR, and nine for AEs. In the ITT population, cabozantinib (surface under the
cumulative ranking curves [SUCRA] 84%), avelumab plus axitinib (SUCRA 68%), and pem-
Consortium
brolizumab plus axitinib (SUCRA 82%) were superior to the other agents for PFS; pem-
Memorial Sloan Kettering Cancer
brolizumab plus axitinib appeared superior for OS (SUCRA 95%); and atezolizumab
Center
demonstrated the lowest likelihood of AEs (SUCRA 100%). Findings were similar in the
First line
intermediate/poor-risk subgroup. Based on the limited data available, avelumab plus
Pembrolizumab axitinib may be preferred in patients with favorable-risk disease.
Avelumab Conclusions: The optimal first-line treatment for mRCC appears to differ by efficacy
endpoint, toxicity, and clinical risk group. Direct comparative studies remain important
in guiding treatment choice.
Patient summary: Head-to-head comparisons do not exist for the newest treatments of
metastatic renal cell carcinoma (mRCC). In an indirect comparison, we found that pem-
brolizumab plus axitinib and cabozantinib are good options for most patients with mRCC.
© 2019 Published by Elsevier B.V. on behalf of European Association of Urology.

y
Equal contribution.
* Corresponding authors. Kidney Cancer Program, City of Hope Comprehensive Cancer Center, 1500 E.
Duarte Road, Duarte, CA 91010, USA. Tel. +1 626 256 4673. E-mail address: spal@coh.org (S.K. Pal);
Division of Urology, Department of Surgery, University of Toronto, 149 College Street, Room 503G,
Toronto, ON M5T 1P5, Canada. Tel: 416-770-2960. E-Mail: wallis.cjd@gmail.com (J.D. Wallis, MD PhD).

https://doi.org/10.1016/j.euo.2019.09.002
2588-9311/© 2019 Published by Elsevier B.V. on behalf of European Association of Urology.
 E U R O P E A N U R O L O GY O N C O L O GY 2 ( 2 019 ) 7 0 8 – 715
1. Introduction
The sequential development of treatments that target
angiogenesis and modulate the immune system has changed
the treatment landscape for patients with metastatic renal cell
carcinoma (mRCC). Clear cell mRCC is characterized by
alterations to the VHL gene, which results in downstream
activation of angiogenesis factors, such as vascular endothelial
growth factor (VEGF) and hypoxia-inducible factor (HIF)
[1]. Treatments that target angiogenesis were shown to
significantly improve survival in mRCC, and approved
angiogenesis-targeted agents now include sunitinib, pazopa-
nib, bevacizumab, axitinib, cabozantinib, sorafenib, and
lenvatinib [2–7]. Metastatic RCC is also considered an
immunogenic tumor based on its history of spontaneous
regression of metastases in response to localized therapy of
the primary tumor and complete responses to high-dose
interleukin-2 [8,9]. Accordingly, immune checkpoint inhibi-
tors (ICIs), including nivolumab monotherapy and nivolumab
plus ipilimumab, were shown to further improve survival in
mRCC, and are approved for salvage and first-line treatment of
patients with mRCC [10,11]. Investigators have proceeded to
study combinations of VEGF-targeted therapy and ICIs. In the
past year, phase 3 clinical trials of avelumab plus axitinib and
pembrolizumab plus axitinib have demonstrated further
improvements in survival [12,13].
Owing to the rapid development of therapies for mRCC,
few head-to-head comparisons of contemporary first-line
treatments exist. Additionally, genetic biomarkers do not
predict differential response to first-line treatment, and the
role for tumor programmed death-ligand 1 (PD-L1) expression
as a predictive biomarker is unclear in mRCC [14]. The
Memorial Sloan Kettering Cancer Center (MSKCC) and
International mRCC Database Consortium (IMDC) risk scores
are validated prognostic models that utilize readily available
clinical data and have become increasingly important in trial
design and patient selection [15,16]. Ultimately, the variable
incorporation of PD-L1 status and clinical risk scores into
registration trial design, and the lack of head-to-head trials
have made it difficult for clinicians to select first-line
treatment for patients with mRCC. To date, guidelines have
typically been informed by expert opinion rather than
quantitative comparison of available agents. In 2018, we
performed a network meta-analysis of the available agents in
an attempt to guide treatment selection [17,18]. Owing to
newly available data and the uncertainty to optimal treat-
ments, herein we perform an indirect comparison of
contemporary first-line treatments for mRCC, including
subgroup analyses according to clinical risk grouping and
PD-L1 expression.
2. Evidence acquisition
We performed a systematic review and network meta-
analysis of parallel-group randomized controlled trials
(RCTs) comparing two systemic therapies in the first-line
treatment of mRCC. This manuscript provides an update of a
previous analysis, with corresponding methodology
[17]. We conducted and reported this review based on
709
the Preferred Reporting Items for Systematic Reviews and
Meta-Analyses (PRISMA) guidelines [19].
To briefly summarize, English-language publications
reporting parallel-group RCTs assessing the treatment of
patients who had not previously received systemic therapy
for mRCC were included. A “clinically relevant” subset of the
available studies was identified based on content expert
opinion. As tyrosine-kinase inhibitors are widely employed
as standard of care, we excluded studies that included
interferon or placebo as control arms. Further, we excluded
experimental therapies that are not yet available in routine
clinical practice and studies that compared differing
treatment schedules of the same medications. Studies were
included if they reported overall survival (OS), progression-
free survival (PFS), or objective response rates (ORRs). We
additionally assessed the rates of serious adverse events
(AEs; grade 3). Utilizing the same methodology, the
previous literature search was updated as of February 17,
2019. Again, data abstraction was performed in duplicate
and risk of bias assessment was performed using the
Cochrane Collaboration’s tool for assessing the risk of bias.
The available direct comparisons between agents were
summarized using a network diagram for each outcome.
Meta-analysis of OS, PFS, and ORR was conducted in the
context of Bayesian hierarchical log-linear models, with both
within- and between-study variance components. When
assessing PFS and OS, we applied contrast-based analyses
using estimated differences in log hazard ratio and standard
error calculated based on published hazard ratios and
confidence intervals using the method of Woods et al
[20]. The relative treatment effects were presented as hazard
ratio and 95% credible interval (CrI). For the assessment of AEs,
we applied arm-based analyses to estimate the odds ratios of
the AEs (and 95% CrI) using the available raw data presented in
the manuscripts. We also estimated the relative ranking of the
different treatments for each outcome using the distribution
of the ranking probabilities and the surface under the
cumulative ranking curves (SUCRA) [21]. Heterogeneity was
assessed using I2 where more than one trial was available for a
given comparison.
Primary analyses for each outcome were performed using
the intention-to-treat (ITT) population of each trial. Prespe-
cified subgroup analyses were performed in the following
groups: intermediate/poor-risk disease and favorable-risk
disease (defined according to MSKCC or IMDC risk categori-
zation at the discretion of individual investigators). Analyses
described were conducted in a Bayesian framework using
“GeMtc” package in R 3.4.2 (https://www.r-project.org/) [22].
The Institutional Review Boards of the University of Utah
and City of Hope Comprehensive Cancer Center approved
this study, and all patients provided informed consent for
collection of clinical data.
3. Evidence synthesis
3.1. Literature search results
Our initial literature search is reported in detail in a prior
publication [17], and an updated search was completed on
710 E U RO P E A N U R O L O GY O N C O L O GY 2 ( 2 019 ) 7 0 8 – 715
Fig. 1 – PRISMA diagram.
PRISMA = Preferred Reporting Items for Systematic Reviews and Meta-Analyses.
a
Full literature search available [17].
February 17, 2019. A total of 502 additional titles were
reviewed, and subsequently, two phase 3 trials were added
to this updated analysis (Fig. 1), for a total of 12 trials in the
quantitative analysis [12,13].
3.2. Characteristics of enrolled trials
All studies enrolled patients within the past decade,
including nine using sunitinib as the control arm [6,11–
13,23–27]. Nine trials reported OS outcomes [6,11–
13,23,24,28–30], 12 reported PFS outcomes [6,11–13,23–
30], 10 reported ORR outcomes [6,11–13,23–25,27–29], and
nine reported grade 3/4 AEs [6,11–13,23,25,27,28,30]
. Furthermore, seven studies reported outcomes stratified
by intermediate/poor-risk patients [6,11–13,28,29], includ-
ing three for OS, seven for PFS, and three for ORR; three
studies reported outcomes by favorable risk stratification
[11–13]. Essentially all trials included patients with a
median age in the 60 s, and the percentage of men in the
majority of studies was 70–75%. Table 1 summarizes the
patient population, treatment arms, and reported efficacy of
12 trials included in the ITT network meta-analysis.
As two studies included data in abstract form only
[23,26], we were unable to assess their risk of bias. Of the
remaining 10 studies, all were open-label trials, with four
trials blinding radiologists assessing progression of disease
[12,13,28,29]. We felt that all studies were at low risk of
attrition and reporting bias (Table 2).
3.3. ITT population
A network meta-analysis of 12 differing agents was
performed for the primary outcome of PFS. Compared with
sunitinib, atezolizumab plus bevacizumab, avelumab plus
axitinib, cabozantinib, nivolumab plus ipilimumab, and
pembrolizumab plus axitinib demonstrated significantly
improved PFS (Table 3 and Fig. 2A). Based on Bayesian
analysis and analysis of the SUCRA, cabozantinib had the
 E U R O P E A N U R O L O GY O N C O L O GY 2 ( 2 019 ) 7 0 8 – 715 711

Table 1 – Summary of clinical trials included in the first-line, intention-to-treat network meta-analysis.

Study Population Arms OS PFS ORR (%) Grade 3/4

AEs (%)

CheckMate 214 ITT, all risk, and PD-L1 status Nivolumab + ipilimumab 0.71 (0.59–0.86) 0.85 (0.73–0.98) 41 45
(N = 550)
Sunitinib (N = 546) 34 61
CABOSUN ITT, intermediate/poor-risk only Cabozantinib (N = 79) 0.80 (0.53–1.21) 0.66 (0.46–0.95) 33 67
Sunitinib (N = 78) 12 60
IMmotion 151 ITT by IRC, all risk groups, Atezolizumab + bevacizumab 0.81 (0.63–1.03) 0.88 (0.74–1.04) 33 40
and PD-L1 status (N = 454)
Sunitinib (N = 461) 31 54
JAVELIN Renal 101 ITT, all risk, and PD-L1 status Avelumab + axitinib (N = 442) 0.78 (0.55–1.08) 0.69 (0.56–0.84) 51 71
Sunitinib (N = 444) 26 72
KEYNOTE-426 ITT, all risk, and PD-L1 status Pembrolizumab + axitinib 0.53 (0.38–0.74) 0.69 (0.57–0.84) 59 76
(N = 432)
Sunitinib (N = 429) 36 71
COMPARZ ITT, all risk groups Pazopanib (N = 557) 0.91 (0.76–1.08) 1.05 (0.90–1.22) 31 NA
Sunitinib (N = 553) 24
Hutson Lancet ITT, all risk groups Axitinib (N = 192) 1.0 (0.73–1.36) 0.77 (0.56–1.05) 32 33
Onc 2013 Sorafenib (N = 96) 15 25
Motzer JCO 2013 ITT, all risk groups Tivozanib (N = 260) 1.25 (0.95–1.62) 0.80 (0.64–0.99) 33 NA
Sorafenib (N = 257) 23
SWITCH ITT, all risk groups Sorafenib (N = 182) NA 1.19 (0.97–1.47) 31 64
Sunitinib (N = 183) 29 65
CROSS-J-RCC ITT, all risk groups Sorafenib (N = 63) NA 0.67 (0.42–1.08) NA NA
Sunitinib (N = 57)
ROPETAR ITT, all risk groups Pazopanib alt. everolimus 0.90 (0.51–1.58) 0.81 (0.50–1.29) NA 42
(N = 52)
Pazopanib (N = 49) 49
IMmotion 150 All risk groups and PD-L1 status Atezolizumab (N = 103) NA 1.19 (0.82–1.71) 25 40
Sunitinib (N = 101) 29 69

AEs = adverse events; IRC = independent review committee; ITT = intention to treat; NA = not available; ORR = objective response rate; OS = overall survival; PD-
L1= programmed death-ligand 1; PFS = progression-free survival.

highest likelihood of providing the maximal PFS (posterior
probability of #1 ranking = 38%, SUCRA 84%). Avelumab plus
axitinib and pembrolizumab plus axitinib were similarly
likely to be deemed the preferred treatment choice (20%
and 17%, respectively; Table 3). Heterogeneity was low in
this analysis with a global I2 of 11.2%.
In the ITT population, a network meta-analysis of eight
agents was performed for the outcome of OS. Compared
with sunitinib, nivolumab plus ipilimumab and pembroli-
zumab plus axitinib demonstrated significantly improved
OS (Table 3 and Fig. 2C). Based on Bayesian analysis and
analysis of the SUCRA, it was highly likely that pembroli-
zumab plus axitinib was the preferred treatment option
(posterior probability of #1 ranking = 77%, SUCRA 95%;
Table 3). Heterogeneity was low in this analysis with a
global I2 of 0%.
A network meta-analysis of nine agents was performed for
the outcome of serious (grade 3) AEs. Compared with
sunitinib, atezolizumab monotherapy, atezolizumab plus
bevacizumab, and nivolumab plus ipilimumab demonstrated

Table 2 – Assessment of risk of bias according to the Cochrane Risk of Bias tool for studies included in the network meta-analysis.

Random sequence Allocation Blinding of Blinding of Incomplete Selective reporting

generation (selection concealment participants and outcome assessment outcome data (reporting bias)
bias) (selection bias) personnel (detection bias) (attrition bias)
(performance bias)

Hutson et al Low risk Unclear At risk Low risk Low risk Low risk
COMPARZ Low risk Unclear At risk At risk Low risk Low risk
Motzer et al Low risk Unclear At risk Low risk Low risk Low risk
SWITCH Low risk Unclear At risk At risk Low risk Low risk
ROPETAR Low risk Unclear At risk At risk Low risk Low risk
IMmotion 150 Low risk Unclear At risk At risk Low risk Low risk
CROSS-J-RCC NA NA NA NA NA NA
CABOSUN Low risk Unclear At risk At risk Low risk Low risk
CheckMate 214 Low risk Unclear At risk At risk Low risk Low risk
IMmotion 151 NA NA NA NA NA NA
KEYNOTE-426 Low risk Unclear At risk Low risk Low risk Low risk
JAVELIN renal 101 Low risk Unclear At risk Low risk Low risk Low risk

NA = not accessible because it is a conference abstract.

712 E U RO P E A N U R O L O GY O N C O L O GY 2 ( 2 019 ) 7 0 8 – 715

statistically lower rates of AEs (Table 3 and Fig. 2E). Based on

Posterior probability

AEs = adverse events; HR = hazard ratio; mRCC = metastatic renal cell carcinoma; NA = not available; OR = odds ratio; ORR = objective response rate; OS = overall survival; PFS = progression-free survival; SUCRA = surface
therapy is best
(SUCRA),% (%)a
Bayesian analysis and analysis of the SUCRA, it was highly
likely that atezolizumab had the lowest rate of serious AEs
(posterior probability of #1 ranking = 99%, SUCRA 100%;
Grade 3/4 AEs

99 (100)
Table 3).
1 (86)

0 (46)

0 (45)
0 (77)
0 (21)

0 (13)
0 (18)
Among 10 agents, cabozantinib, avelumab plus axitinib,

NA

NA

NA
pembrolizumab plus axitinib, and axitinib demonstrated
improved ORR compared with sunitinib (Table 3). Based on
OR vs sunitinib

0.52 (0.41–0.67)

0.60 (0.47–0.76)

0.26 (0.16–0.43)

0.99 (0.64–1.5)
0.99 (0.73–1.3)

1.5 (0.75–3.0)
1.3 (0.96–1.8)
Bayesian analysis and analysis of the SUCRA, it is less clear
1.3 (0.7–2.6)

what the superior treatment is for ORR, but avelumab plus

axitinib, cabozantinib, and pembrolizumab plus axitinib
NA

NA

NA
performed better than other comparators (Table 3).
Posterior probability

3.4. Intermediate/poor-risk subgroup

(SUCRA), % (%)a
therapy is best

A network meta-analysis of seven trials was performed to

assess the outcome of PFS among patients with intermedi-
<1 (25)
<1 (45)
<1 (41)

<1 (13)
42 (83)

28 (80)
21 (86)

ate/poor-risk disease. The results of these studies are

7 (78)

2 (59)
ORR

NA
NA

summarized in Supplementary Table 1. The results of the

network meta-analysis in this subgroup are similar to the
OR vs sunitinib

2.94 (2.03–4.26)

1.82 (0.84–3.88)
1.42 (0.99–2.04)

2.54 (1.73–3.64)
0.86 (0.40–1.91)
1.35 (0.97–1.87)

results of the ITT population, with cabozantinib (posterior

2.79 (1.14–7.35)
1.11 (0.76–1.64)
3.2 (1.08–8.89)

probability of #1 ranking = 37%, SUCRA 77%), pembrolizu-

mab plus axitinib (posterior probability of #1 ranking = 27%,
SUCRA 77%), and avelumab plus axitinib (posterior proba-
Table 3 – Efficacy summaries for the intention-to-treat population in first-line mRCC as compared with sunitinib.

NA
NA

bility of #1 ranking = 17%, SUCRA 73%) likely to provide

superior results to alternative therapies (Table 4 and
Posterior probability

Fig. 2B).
(SUCRA), % (%)a
therapy is best

In the intermediate/poor-risk group, the results of the

network meta-analysis for OS are similar to those of the ITT
population, with pembrolizumab plus axitinib highly likely
38 (84)

20 (83)

10 (53)
17 (82)
0 (55)

0 (56)

8 (59)

7 (61)
0 (18)

0 (17)
0 (8)

to be the preferred treatment option (posterior probability

PFS

of #1 ranking = 89%, SUCRA 97%; Table 4 and Fig. 2D). For

Respective OS, PFS, ORR, and grade 3/4 AE SUCRA values for sunitinib are 8%, 26%, 15%, and 45%.

ORR, there is no therapy that is clearly superior to its

HR vs sunitinib

0.66 (0.46–0.96)

0.68 (0.56–0.83)
0.83 (0.72–0.95)

0.69 (0.57–0.84)
0.83 (0.74–0.94)

0.80 (0.55–1.20)
0.85 (0.51–1.40)

0.79 (0.56–1.10)
1.10 (0.90–1.20)

1.2 (0.90–1.60)

1.0 (0.84–1.30)

comparators, although cabozantinib has the highest prob-

ability of being the preferred choice (53%; Table 4).

3.5. Favorable-risk subgroup

Posterior probability

For patients with favorable-risk mRCC, limited data on

(SUCRA), % (%)a
therapy is best

outcomes are reported from clinical trials (Supplementary

Table 2). Of the available results, there is evidence that
avelumab plus axitinib may be superior to its comparators
77 (95)

for PFS in favorable-risk patients (posterior probability of

4 (48)
1 (49)
0 (28)
3 (52)

9 (45)
6 (76)

#1 ranking >99%, SUCRA 100%; Supplementary Table 3).

NA
NA

NA
NA
OS

HR vs sunitinib

3.6. Discussion
(0.58–0.78)

(0.38–0.74)

(0.46–1.5)
(0.53–1.2)
(0.62–1.0)

(0.56–1.1)
(0.8–1.0)

The first-line treatment landscape for mRCC is rapidly

evolving; yet, there is a paucity of head-to-head clinical
0.78
0.53

0.82
0.80
0.80
0.67

0.91

under the cumulative ranking curve.

NA
NA

NA
NA

trials or predictive biomarkers to guide treatment selection.

We have previously performed a network meta-analysis of
Atezolizumab + bevacizumab

these agents in an attempt to guide treatment decisions

Pembrolizumab + axitinib
Nivolumab + Ipilimumab

Pazopanib + everolimus

quantitatively [17]. Since that time, two new trials have

Avelumab + axitinib

been published, which question the validity of prior

findings [12,13]. In this updated, indirect comparison of
Atezolizumab
Cabozantinib

first-line treatments for mRCC, cabozantinib, avelumab plus

Pazopanib

Tivozanib
Sorafenib
Therapy

axitinib, and pembrolizumab plus axitinib are likely to be

Axitinib

the preferred agents with respect to PFS, while pembroli-

zumab plus axitinib may be best for OS in the ITT and
a
 E U R O P E A N U R O L O GY O N C O L O GY 2 ( 2 019 ) 7 0 8 – 715
Fig. 2 – Forest plot assessing systemic therapies, compared with sunitinib. Progression-free survival is assessed in the (A) intention-to-treat and (B)
intermediate/poor-risk populations. Similarly, overall survival is assessed in the (C) intention-to-treat and (D) intermediate/poor-risk populations.
Finally, (E) adverse events are assessed in the intention-to-treat population.
% Crl = 95% credible interval.
intermediate/poor-risk populations. In the same popula-
tion, no single treatment appears to be superior to its
comparators for ORR. Atezolizumab offers the most
favorable profile with respect to serious AEs. Compared
with previous analyses, this work provides updated results
including novel combinations of molecularly targeted
agents and immunotherapy, as well as results stratified
by clinical risk category.
The IMDC and MSKCC risk scores use readily available
clinical factors, including time from diagnosis to systemic
therapy, Karnofsky performance status, hemoglobin, calci-
um, absolute neutrophil count, platelets, and lactate
dehydrogenase, to prognosticate survival at the initiation
of VEGF-targeted therapies and ICIs [15,16,31]. CheckMate
214 showed that nivolumab plus ipilimumab improves
survival in intermediate- and poor-risk patients, but
sunitinib was a better option for patients with favorable-
risk disease [11]. Since CheckMate 214, survival by risk
group is reported in first-line clinical trials, so we performed
a separate meta-analysis for patients with intermediate/
poor-risk and favorable-risk disease. The meta-analysis for
intermediate/poor-risk patients had similar findings across
all efficacy endpoints to the meta-analysis for ITT popula-
tion. The similar results are not surprising as the majority of
patients enrolled in contemporary clinical trials are at an
intermediate or poor risk. The favorable-risk meta-analysis
was limited because CABOSUN did not include favorable-
risk patients, and trials were inconsistent in reporting
713
efficacy for those with favorable-risk disease. With these
limitations in mind, avelumab plus axitinib appeared to be
the best option for patients with favorable-risk disease.
Many clinicians speculate that cabozantinib will have
similar efficacy across IMDC risk groups; yet due to the
lack of trial data, cabozantinib was not included in the
favorable-risk meta-analysis. Finally, due to the unique
mechanism of action of ICIs, the combination of nivolumab
plus ipilimumab was expected to perform poorly in the PFS
meta-analyses.
Owing to the nature of network meta-analyses, our study
has multiple limitations. First, while we compared the
included clinical trials in the ITT population and in multiple
subgroups, the population for the primary endpoint differed
across many of the trials, so a trial may not have been
adequately powered to detect differences in distinct
populations. Second, many clinical trials did not report all
outcomes across clinical risk groups. Third, as discussed in
depth in our previous network meta-analysis, the results of
some of the included trials (particularly with respect to
cabozantinib and tivozanib) are controversial [17]. Fourth,
we relied upon aggregate published data and did not
perform individual patient data meta-analysis. This limits
our ability to assess effect modification. Fifth, as with all
studies relying upon data derived from RCTs, these data are
subject to the efficacy-effectiveness gap and may not be
entirely generalizable to real-world practice. Sixth, the OS
data from KEYNOTE-426, JAVELIN renal 101, and IMmotion
714 E U RO P E A N U R O L O GY O N C O L O GY 2 ( 2 019 ) 7 0 8 – 715

151 are immature, so the findings could change with the

Posterior probability

AEs = adverse events; HR = hazard ratio; mRCC = metastatic renal cell carcinoma; NA = not available; OR = odds ratio; ORR = objective response rate; OS = overall survival; PFS = progression-free survival; SUCRA = surface
(SUCRA), % (%)a
therapy is best
final analyses. Finally, the majority of these studies were
limited to patients with clear cell histology, and thus, it is
unclear to what extent these results are generalizable to
patients with other histologies. Ultimately, we performed
38 (75)

53 (71)
9 (52)
this network meta-analysis in an attempt to aid in clinical
NA

NA
NA
NA
decision making as first-line trials are being rapidly
ORR

reported, but the meta-analysis is not intended to replace

the need for head-to-head clinical trials of contemporary
first-line therapies.
OR vs sunitinib

2.38 (1.60–3.48)

2.52 (0.81–7.37)
1.94 (1.39–2.70)

4. Conclusions
NA

NA
NA
NA

Based on this updated network meta-analysis, the optimal

first-line treatment for mRCC appears to differ by clinical
risk group and efficacy endpoint. For the ITT and interme-
diate/poor-risk groups, cabozantinib, avelumab plus axiti-
Posterior probability

nib, and pembrolizumab plus axitinib are likely to be the

(SUCRA), % (%)a
therapy is best

preferred agents for PFS, while pembrolizumab plus axitinib

may be superior for OS. These findings suggest that
clinicians should take into account IMDC or MSKCC risk
27 (77)
37 (77)
11 (44)
17 (73)

group, risk for toxicities, and patient preference to tailor

2 (54)

6 (47)
0 (14)
Table 4 – Efficacy summaries for the intermediate/poor-risk population in first-line mRCC as compared with sunitinib.

first-line treatment selection for patients with mRCC. Direct

comparative studies remain integral in guiding treatment
PFS

choice, and these results may help inform the design of

future trials.
HR vs sunitinib

0.66 (0.46–0.96)
0.69 (0.56–0.86)
0.67 (0.53–0.85)
0.76 (0.65–0.89)

0.84 (0.49–1.5)

0.83 (0.54–1.3)

Author contributions: Sumanta K. Pal had full access to all the data in the
1.0 (0.73–1.4)

study and takes responsibility for the integrity of the data and the
accuracy of the data analysis.

Study concept and design: Hahn, Klaassen, Agarwal, Pal, Wallis.

Acquisition of data: Hahn, Klaassen, Agarwal, Esther, Pal, Wallis.
Analysis and interpretation of data: Hahn, Klaassen, Agarwal, Haaland, Ye,
Wang, Pal, Wallis.
Drafting of the manuscript: Hahn, Klaassen, Agarwal, Haaland, Esther, Ye,
Posterior probability

(SUCRA), % (%)a
therapy is best

Pal, Wallis.
Critical revision of the manuscript for important intellectual content: Hahn,
Klaassen, Agarwal, Haaland, Esther, Ye, Wang, Pal, Wallis.
Respective OS, PFS, and ORR SUCRA values for sunitinib are 6%, 3%, and 2%.

Statistical analysis: Ye, Haaland, Wang.

89 (97)
3 (44)
8 (73)

Obtaining funding: None.

NA

NA
NA
NA

Administrative, technical, or material support: None.

Supervision: Pal, Agarwal, Wallis.
OS

Other: None.

Financial disclosures: Sumanta K. Pal certifies that all conflicts of

HR vs sunitinib

0.66 (0.54–0.80)

0.50 (0.35–0.71)
0.80 (0.53–1.2)

interest, including specific financial interests and relationships and

affiliations relevant to the subject matter or materials discussed in the
manuscript (eg, employment/affiliation, grants or funding, consultan-
cies, honoraria, stock ownership or options, expert testimony, royalties,
NA

NA
NA
NA

or patents filed, received, or pending), are the following: Andrew W.

under the cumulative ranking curve.

Hahn, Zachary Klaassen, Benjamin Haaland, John Esther, Xiang Y. Ye,

Xuechen Wang, and Christopher J.D. Wallis have no conflicts to report.
Neeraj Agarwal reports consultancy to the following: Pfizer, Novartis,
Pembrolizumab + axitinib

Merck, Genentech, Eisai, Exelixis, Clovis, EMD Serono, BMS, Astra Zeneca,
Nivolumab + Ipilimumab

Foundation One, Astellas, Ely Lilly, Bayer, Argos, Medivation, Clovis, and
Avelumab + axitinib

Nektar. Neeraj Agarwal also reports research funding to Sumanta K. Pal’s

institution on Sumanta K. Pal’s behalf from the following companies:
Cabozantinib

Active Biotech, Astra Zeneca, Bavarian Nordic, BMS, Calithera, Celldex,

Tivozanib
Sorafenib
Therapy

Axitinib

Eisai, Exelixis, Genetech, GSK (GlaxoSmithKline), Immunomedics,

Janssen, Medivation, Merck, NewLink Genetics, Novartis, Pfizer, Prome-
theus, Rexahn, Sanofi, Takeda, and Tracon. Sumanta K. Pal reports
a
 E U R O P E A N U R O L O GY O N C O L O GY 2 ( 2 019 ) 7 0 8 – 715
consultancy to the following: Genentech, Aveo, Eisai, Roche, Pfizer,
Novartis, Exelixis, Ipsen, BMS, and Astellas.
Funding/Support and role of the sponsor: None.
Appendix A. Supplementary data
Supplementary material related to this article can be
found, in the online version, at doi:https://doi.org/10.1016/j.
euo.2019.09.002.
References
[1] Rini BI, Campbell SC, Escudier B. Renal cell carcinoma. Lancet
2009;373:1119–32.
[2] Motzer RJ, Hutson TE, Tomczak P, et al. Sunitinib versus interferon
alfa in metastatic renal-cell carcinoma. N Engl J Med
2007;356:115–24.
[3] Sternberg CN, Davis ID, Mardiak J, et al. Pazopanib in locally
advanced or metastatic renal cell carcinoma: results of a random-
ized phase III trial. J Clin Oncol 2010;28:1061–8.
[4] Escudier B, Bellmunt J, Negrier S, et al. Phase III trial of bevacizumab
plus interferon alfa-2a in patients with metastatic renal cell carci-
noma (AVOREN): final analysis of overall survival. J Clin Oncol
2010;28:2144–50.
[5] Rini BI, Escudier B, Tomczak P, et al. Comparative effectiveness of
axitinib versus sorafenib in advanced renal cell carcinoma (AXIS): a
randomised phase 3 trial. Lancet 2011;378:1931–9.
[6] Choueiri TK, Halabi S, Sanford BL, et al. Cabozantinib versus suni-
tinib as initial targeted therapy for patients with metastatic renal
cell carcinoma of poor or intermediate risk: the alliance A031203
CABOSUN trial. J Clin Oncol 2017;35:591–7.
[7] Motzer RJ, Hutson TE, Glen H, et al. Lenvatinib, everolimus, and the
combination in patients with metastatic renal cell carcinoma: a
randomised, phase 2, open-label, multicentre trial. Lancet Oncol
2015;16:1473–82.
[8] Rosenberg SA, Lotze MT, Muul LM, et al. A progress report on the
treatment of 157 patients with advanced cancer using lymphokine-
activated killer cells and interleukin-2 or high-dose interleukin-2
alone. N Engl J Med 1987;316:889–97.
[9] De Meerleer G, Khoo V, Escudier B, et al. Radiotherapy for renal-cell
carcinoma. Lancet Oncol 2014;15:e170–7.
[10] Motzer RJ, Escudier B, McDermott DF, et al. Nivolumab versus
everolimus in advanced renal-cell carcinoma. N Engl J Med
2015;373:1803–13.
[11] Motzer RJ, Tannir NM, McDermott DF, et al. Nivolumab plus ipili-
mumab versus sunitinib in advanced renal-cell carcinoma. N Engl J
Med 2018;378:1277–90.
[12] Rini BI, Plimack ER, Stus V, et al. Pembrolizumab plus axitinib versus
sunitinib for advanced renal-cell carcinoma. N Engl J Med
2019;380:1116–27.
[13] Motzer RJ, Penkov K, Haanen J, et al. Avelumab plus axitinib versus
sunitinib for advanced renal-cell carcinoma. N Engl J Med
2019;380:1103–15.
[14] Zhu J, Armstrong AJ, Friedlander TW, et al. Biomarkers of immuno-
therapy in urothelial and renal cell carcinoma: PD-L1, tumor muta-
tional burden, and beyond. J Immunother Cancer 2018;6:4.
715
[15] Motzer RJ, Mazumdar M, Bacik J, Berg W, Amsterdam A, Ferrara J.
Survival and prognostic stratification of 670 patients with advanced
renal cell carcinoma. J Clin Oncol 1999;17:2530–40.
[16] Ko JJ, Xie W, Kroeger N, et al. The International Metastatic Renal Cell
Carcinoma Database Consortium model as a prognostic tool in
patients with metastatic renal cell carcinoma previously treated
with first-line targeted therapy: a population-based study. Lancet
Oncol 2015;16:293–300.
[17] Wallis CJD, Klaassen Z, Bhindi B, et al. First-line systemic therapy for
metastatic renal cell carcinoma: a systematic review and network
meta-analysis. Eur Urol 2018;74:309–21.
[18] Hahn AW, Hale P, Maughan BL, Haaland B, Agarwal N. Optimal first-
line treatment of metastatic renal-cell carcinoma: a network meta-
analysis. Kidney Cancer 2018;2:115–21.
[19] Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items
for systematic reviews and meta-analyses: the PRISMA statement. J
Clin Epidemiol 2009;62:1006–12.
[20] Woods BS, Hawkins N, Scott DA. Network meta-analysis on the log-
hazard scale, combining count and hazard ratio statistics account-
ing for multi-arm trials: a tutorial. BMC Med Res Methodol
2010;10:54.
[21] Salanti G, Ades AE, Ioannidis JP. Graphical methods and numerical
summaries for presenting results from multiple-treatment meta-
analysis: an overview and tutorial. J Clin Epidemiol 2011;64:163–71.
[22] Neupane B, Richer D, Bonner AJ, Kibret T, Beyene J. Network meta-
analysis using R: a review of currently available automated
packages. PLoS One 2014;9:e115065.
[23] Motzer RJ, Powles T, Atkins MB, et al. IMmotion151: a randomized phase
III study of atezolizumab plus bevacizumab vs sunitinib in untreated
metastatic renal cell carcinoma (mRCC). J Clin Oncol 2018;36:578.
[24] Motzer RJ, Hutson TE, Cella D, et al. Pazopanib versus sunitinib in
metastatic renal-cell carcinoma. N Engl J Med 2013;369:722–31.
[25] Eichelberg C, Vervenne WL, De Santis M, et al. SWITCH: a random-
ised, sequential, open-label study to evaluate the efficacy and safety
of sorafenib-sunitinib versus sunitinib-sorafenib in the treatment
of metastatic renal cell cancer. Eur Urol 2015;68:837–47.
[26] Tomita Y, Naito S, Sassa N, et al. Sunitinib versus sorafenib as first-line
therapy followed by sorafenib and sunitinib for patients with meta-
static renal cell carcinoma (RCC) with clear cell histology: a multi-
center randomized trial, CROSS-J-RCC. J Clin Oncol 2017;35:469.
[27] McDermott DF, Huseni MA, Atkins MB, et al. Clinical activity and
molecular correlates of response to atezolizumab alone or in com-
bination with bevacizumab versus sunitinib in renal cell carcinoma.
Nat Med 2018;24:749–57.
[28] Hutson TE, Lesovoy V, Al-Shukri S, et al. Axitinib versus sorafenib as
first-line therapy in patients with metastatic renal-cell carcinoma: a
randomised open-label phase 3 trial. Lancet Oncol 2013;14:1287–94.
[29] Motzer RJ, Nosov D, Eisen T, et al. Tivozanib versus sorafenib as initial
targeted therapy for patients with metastatic renal cell carcinoma:
results from a phase III trial. J Clin Oncol 2013;31:3791–9.
[30] Cirkel GA, Hamberg P, Sleijfer S, et al. Alternating treatment with
pazopanib and everolimus vs continuous pazopanib to delay dis-
ease progression in patients with metastatic clear cell renal cell
cancer: the ROPETAR randomized clinical trial. JAMA Oncol
2017;3:501–8.
[31] Yip SM, Wells C, Moreira R, et al. Checkpoint inhibitors in patients
with metastatic renal cell carcinoma: results from the International
Metastatic Renal Cell Carcinoma Database Consortium. Cancer
2018;124:3677–83.