EURURO Vol. 72 No. 3

378 476

Since this trial used RECIST 1.1 and could misclassify
some patients with pseudoprogression, the study allowed
investigators to treat past RECIST progression with patient
re-consent. During the trial, updates to RECIST were
proposed using immune-modified criteria, but were not
incorporated In the trial, 316 of the 406 (78%) patients
treated with nivolumab progressed according to RECIST; of
these, 153 (48%) were treated beyond progression. Among
the patients treated past progression, 48% had some degree
of tumor burden reduction and 13% had a reduction in
tumor size 30% according to RECIST. These data support a
model of delayed response in some patients, and there
appeared to be some relationship between delayed
responses and the pattern of initial response to nivolumab.
For example, delayed responses were observed in 28%, 14%,
and 6% of patients who had an initial partial/complete
response, stable disease, or initial progressive disease,
respectively, as their best initial response.
Survival was improved in patients treated beyond
progression compared to those not treated beyond pro-
gression. However, patients treated beyond progression
were more likely to have good functional status, no
functional deterioration, and a favorable prognostic risk
group. While the authors did not report on treatment with
everolimus beyond progression, we suspect that similar
survival benefits may be observed given differences in
patient selection. Thus, improvements in survival may be
related not only to delayed responses with nivolumab but
also to the more favorable prognosis of patients treated
beyond progression. Only through randomized discontinu-
ation studies of immunotherapy can we be certain about the
optimal timing of therapy cessation.
The decision to stop therapy can be agonizing for
patients and physicians: on one hand, delayed responses,
while rare, are possible; on the other hand, alternative and
effective therapies may be available, including clinical trials.
Escudier et al have attempted to address this, and found
that patients with a favorable Memorial Sloan Kettering
Cancer Center (MSKCC) risk group, lung metastases, and
Karnofsky performance status (KPS) 90 are more likely to
respond. However, responses were still observed in patients
with liver metastases, lower KPS, or intermediate prognosis
according to MSKCC criteria. Clearly, clinical criteria are at
present unable to predict this delayed response, and further
research on biomarkers of immune response are needed.
Interestingly, while RCC patients with poor risk seemed to
have the greatest survival advantage with nivolumab over
everolimus, these patients did not benefit from continued
nivolumab beyond progression.
Because of pseudoprogression, RECIST has fortunately
been further updated for immune therapies, and is now
called iRECIST The main change is that first progression
on scans becomes the new baseline from which to measure
tumor change, and true progression is defined with
confirmation of further progression on subsequent scans
from this new baseline. Future trials incorporating iRECIST
could permit continued therapy as part of the original
design, provided that ongoing clinical benefit is observed.
Like all quantitative metrics however, future trials will need
to examine how new surrogate outcomes fully capture the
survival benefits of immunotherapy.
Given that only a minority of mRCC patients respond to
single-agent nivolumab at any point, further work is needed
to clarify which patients are more likely to benefit and
which systemic therapy combinations and sequences
optimize survival benefits. Until then, the findings reported
by Escudier et al will reassure clinicians that some patients
do benefit from therapy beyond an initial growth in tumor
size, and that it remains important to treat the patient and
not the CT findings.
Conflicts of interest: Andrew J. Armstrong has received institutional
research funding from Medivation, Astellas, Bayer, Dendreon, Janssen,
Active Biotech, Sanofi Aventis, Gilead, Novartis, and Pfizer, consulting/
speaker fees fromDendreon and Sanofi Aventis, and consulting fees from
Medivation, Astellas, and Janssen. Michael S. Humeniuk has nothing to
disclose.
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
.
E U R O P E A N U R O L O G Y 7 2 ( 2 0 1 7 ) 3 7 7 – 3 7 8 378