Multicenter Phase 2 Study to Identify the Optimal neo-Adjuvant Combination Scheme of Ipilimumab and Nivolumab (OpACIN-neo) and Personalized Response-driven Adjuvant Combination (PRADO extension cohort)

T cell checkpoint blockade by anti-CTLA and/or anti-PD1 is currently the most
promising therapy in late stage melanoma to induce long-term benefit or even
cure. Particularly the combination of ipilimumab and nivolumab induces high
response rates and promising response depth. This raises the question whether
ipilimumab and/or nivolumab could also become standard therapy in adjuvant
treatment of melanoma. Ipilimumab as adjuvant therapy significantly improves
recurrence free survival. Recently overall survival data were published and
demonstrated a significant overall survival benefit for patients in the
ipilimumab group with a 5-year overall survival rate of 65% compared to 54% in
the placebo group. Nivolumab and pembrolizumab have been tested in large phase
3 randomized trials showing improved relapse free survival (RFS) compared to
placebo and ipilimumab at excellent tolerability, but overall survival (OS)
data are pending from both trials. In addition, adjuvant targeted therapy by
the BRAF+MEK inhibitor combination of dabrafenib + trametinib in BRAFV600E/K
mutation positive patients has been shown to improve RFS and OS (only RFS
significant).
In contrast to chemotherapeutic approaches, immunotherapeutic approaches depend
on sufficient activation of the immune system. To become fully activated, T
cells require two signals. The first signal is provided by the interaction of
the (tumor-) antigen presented in the major histocompatibility complex (MHC) on
the antigen-presenting cell (APC) to the T cell receptor (TCR) on the T cell
(signal 1). In parallel, a large number of co-inhibitory and co-stimulatory
interactions - so-called T cell checkpoints - modulate the outcome of the TCR -
pMHC interaction. Antibody-based interference with the T cell checkpoints
CTLA-4 and PD-1 has been shown to improve tumor-specific T cell responses and
to result in a significant clinical benefit in patients with melanoma and other
cancers.
The notion that T cell checkpoint inhibition is of greatest value at the moment
of TCR triggering has potentially significant consequences for the use of
checkpoint targeting antibodies as adjuvant therapies. Specifically, as the
amount of antigen that can provide this signal 1 will correlate with tumor
load, adjuvant immunotherapy can be assumed to work most efficiently, when
adjuvant treatment is initiated prior to surgery. A phase 1 study testing
neo-adjuvant and adjuvant application of 4 courses of ipilimumab + nivolumab in
stage III melanoma (OpACIN, CA209-278, NCT02437279) showed that neo-adjuvant
immunotherapy was feasible, but with the current standard regimen (4 courses
ipilimumab 3mg/kg + nivolumab 1mg/kg) to toxic for (neo)adjuvant application.
Pathologic responses after only 6 weeks of neoadjuvant ipilimumab + nivolumab
were high with 78%, and all responders are free of relapse with a median
follow-up of 25 months.
The observed feasibility and pathologic responses, combined with the observed
toxicity, form the rationale for the design of the subsequent phase 2 trial,
OpACIN-neo, that will compare safety and efficacy of neo-adjuvant ipilimumab +
nivolumab of three different schemes combining ipilimumab and nivolumab. Aim is
to identify a combination scheme that maintains the high rate of clinical
responses observed, but at lower toxicity. The interim analysis after 39
patients having undergone the complete lymph node dissection (CLND) confirmed
the high pathologic response rates of more than 60% in all arms, while toxicity
was lower with 38%, 15%, and 50% for arm A, B, and C respectively (unplanned
interim analysis at 62 patients due to one severe toxicity observed in Arm C).
These interim data indicate that it is likely that OpACIN-neo will meet its
primary objective to identify a neoadjuvant combination scheme that is less
toxic, but similar effective. Thus, one of the coprimary objectives of the
PRADO extension cohort will be to confirm the safety and efficacy of the winner
combination. None of the patients with pathologic response in OpACIN and
OpACIN-neo has relapsed so far. Interestingly, in stage IV melanoma long-term
benefit is observed in patients achieving complete response upon nivolumab or
ipilimumab + nivolumab with 88% and 80% PFS at 3 years (Robert et al. ESMO
2017). Even after discontinuation of checkpoint inhibition responses are
ongoing in the vast majority of patients. Recently updated data from the
KEYNOTE 006 trial demonstrated that for patients with CR as best overall
response that stopped pembrolizumab after 2 years, the estimated progression
free survival rate at 18 months after stop of therapy was 96% (Long et al
presented at ASCO 2018). These observations question the benefit of CLND in
patients achieving deep pathologic response after neoadjuvant ipilimumab +
nivolumab. Therefore, additional coprimary objectives of PRADO will be to test
response-driven intensities of subsequent therapy (surgery and adjuvant
immunotherapy) after neoadjuvant ipilimumab + nivolumab.

OpACIN-neo:
Primary objectives:
- To compare safety (immune related adverse events during the first 12 weeks)
of three different neo-adjuvant combination schemes of ipilimumab + nivolumab
- To compare radiological and pathologic response rates at week 6 of the three
neo-adjuvant combination schemes
Secondary objectives:
- To describe relapse free survival
- To describe rate and type of late adverse events (up to 3 years)
- To determine immune activating capacity of the alternative combination
schemes
- Antigen specific T cell responses in peripheral blood from baseline to
surgery at week 6.
PRADO extension cohort:
Primary objectives:
- To confirm the pathologic response rate (index lymph node, at week 6) of the
winner neoadjuvant ipilimumab + nivolumab combination scheme identified in
OpACIN-neo
- To show that patients with pathologic complete response (pCR) or pathologic
near complete response (pnCR) in the index lymph node can be spared safely a
CLND without affecting their prognosis
- To improve RFS at 24 months for patients achieving no pathologic response
(pNR) by intensifying the subsequent adjuvant therapy
Secondary objectives:
- To confirm radiologic response rate at week 6 of the winner neo-adjuvant
combination scheme identified in OpACIN-neo
- To describe RFS and distant metastasis free survival (DMFS), for pCR, pnCR,
pPR, and pNR patients
- To describe event free survival (EFS) for the total population
- To describe OS for the total population and for pCR/pnCR, pPR and pNR patients
- To describe the grade 3/4 immune related adverse event rate within the first
12 weeks
- To describe rate and type of late adverse events (up to 3 years)
- To determine differences in surgical morbidity for only marked index lymph
node resection versus CLND.
- To explore the immune activating capacity of the different response-driven
treatment schemes
- To identify baseline biomarkers predictive for pCR/pnCR
- To evaluate health related quality of life

This is an open-label three-arm phase 2 trial (including a Simon stage 2
design) consisting of 90 stage III melanoma patients randomized 1:1:1 to
receive either 2 courses 3 mg/kg ipilimumab + 1 mg/kg nivolumab every 3 weeks
(Arm A), 2 courses 1 mg/kg ipilimumab + 3 mg/kg nivolumab every 3 weeks (Arm
B), or 2 courses ipilimumab 3 mg/kg, directly followed by 2 courses nivolumab 3
mg/kg every 2 weeks (Arm C). All three treatment arms are applied prior to
surgery at week 6, and are neo-adjuvant (30 patients per arm, Figure 1).
Patients will be stratified according to treatment center. An interim analysis
will be performed after 13 patients have been included in each arm (see 4.5),
thus in total 39 patients have been included. After inclusion of the 90
patients OpACIN-neo will transiently closed for analysis of the toxicity and
response rate, and together with DSMB and BMS the study team will pick the
winner scheme. Subsequent the extension cohort (PRADO) will be opened to
confirm the response rates observed from the winner scheme in OpACIN-neo and to
analyze the requirement of CLND after achieving pCR or pnCR in the index lymph
node.
The extension cohort will consist of about 100-110 patients (inclusion will
stop when 50 patients have achieved pCR or pnCR in their index lymph node). All
patients in the extension cohort will receive (after marker placement into the
largest lymph node metastasis) the same neoadjuvant ipilimumab + nivolumab
combination scheme (the OpACIN-neo winner scheme). After 6 weeks of treatment,
the patients will undergo a solitary lymph node removal of the marked index
lymph node. Patients achieving pCR or pnCR will not undergo a subsequent CLND
and will not receive any adjuvant therapy. Patients achieving pPR will undergo
CLND and start follow-up thereafter in line with OpACIN-neo without any
adjuvant therapy. Patients achieving pNR will undergo CLND and start at week 12
adjuvant radiotherapy (at patient*s and treating physician*s discretions) and
adjuvant nivolumab 480mg q4wks for 52 weeks. In case of commercially available
pNR patients that are BRAF V600E/K mutation positive may also receive adjuvant
dabrafenib + trametinib (then after radiotherapy when indicated) instead of
nivolumab according to the patient*s and the treating physician*s decision

In OpACIN-neo patients will be treated pre-surgically for 6 weeks with the
combination of ipilimumab + nivolumab at three different combination schemes
(30 patients per arm).
Medicine tested: 2 courses ipilimumab 3 mg/kg plus nivolumab 1 mg/kg q3wks,
versus 2 courses ipilimumab 1 mg/kg plus nivolumab 3 mg/kg q3wks, versus 2
courses ipilimumab 3 mg/kg q3wks, directly followed by 2 courses nivolumab 3
mg/kg every 2 weeks.
Lab testing (incl. collection of PBMC and EDTA blood for plasma and thrombocyte
isolation) will be performed during screening, at baseline, week 6 pre-surgery
and directly post-surgery, and at week 12.
Tumor biopsies/material preservation is required at baseline and during
surgery.
CT scans will be required at baseline and at week 6.
Follow-up will start at week 12 with a CT or PET /CT scan according to the
individual center*s standard. Subsequent follow-up will be by PET/CT or CT
scans according to the national/institutional standards for high-risk melanoma
(CT 3 monthly in the first years in the extension cohort).
Collection of another PBMC and EDTA blood (for plasma and thrombocyte
isolation) sample, and collection of tumor biopsies will be performed at the
time point of relapse.
In the PRADO extension cohort patients will be treated pre-surgically for 6
weeks with the winner combination of ipilimumab + nivolumab (about 100-110
patient, stopping inclusion when 50 patients have achieved pCR or pnCR).

Until recently, there was no standard adjuvant therapy for patients with
high-risk stage III melanoma. Post-surgery adjuvant radiotherapy is commonly
applied, because it has been shown to marginally improve local disease control,
but neither benefit in RFS, nor overall survival (OS), can be achieved in this
high-risk patient population [1, 2]. High-dose interferon is currently the only
systemic therapy for the adjuvant treatment of melanoma that is approved in
some countries. However, questionable survival benefit (two out of three
meta-analyses found no OS benefit [3-5]) and serious toxicity has led to the
fact that interferon is not a generally accepted adjuvant treatment.
Recently, adjuvant systemic targeted therapy with dabrafenib and trametinib in
BRAFV600E/K mutation positive patients has been shown to improve significant
RFS and OS (not significant) [6]. Grade 3/4 toxicities were reported to be 41%.
Adjuvant systemic immunotherapy with ipilimumab is the only treatment so far to
improve RFS and OS in stage III melanoma [7]. Immune-related grade 3 or 4
adverse events occurred in 41.6% of patients treated with ipilimumab 10 mg/kg
adjuvant. Due to the high grade of toxicity and due to the upcoming trial with
anti-PD-1adjuvant ipilimumab was not filed for approval as adjuvant therapy.
Recently, two phase 3 trials showed RFS benefit from anti-PD-1 adjuvant. Both,
nivolumab and pembrolizumab improved significantly RFS (as compare to
ipilimumab and placebo respectively), however OS benefit data are pending.
Treatment-related grade 3/4 toxicity rates were clearly lower with 14.4% and
14.7% [8, 9]. Screen failure rate was high in the adjuvant pembrolizumab E1325
trial, with a total of 30% screen failures, and 14% due to early relapse within
12 weeks post-surgery, making RFS judgement on adjuvant and neoadjuvant trials
(latter representing more intention to treat populations) hard to compare due
to the high patient selection in the adjuvant trials. Due to the lack of OS
data, adjuvant PD-1 blockade might be delayed in approval and availability for
stage III melanoma patients.
Participants of this study will be exposed to two immunotherapeutic agents
(ipilimumab and nivolumab) known to induce immune related adverse events at a
high percentage when combined together [10]. This has been also observed in the
precursor phase 1 trial (OpACIN) in which 90% of the patients developed grade
3/4 toxicities (Blank et al., in revision). 4 patients developed severe irAE:
one patient developed life-threatening Steven-Johnson syndrome (after 3 courses
ipilimumab + nivolumab) that recovered upon mycophenolate, one patient
developed severe long-term colitis that has been stabilized with tacrolimus and
recovered after months, one patient developed insulin dependent diabetes
mellitus, and one patient developed polyradiculitis recovering after 1 year of
revalidation. Due to these toxicities all, except 2 patients, had to stop
earlier resulting in a maximum application of 2-3 courses of ipilimumab +
nivolumab, independent of neo-adjuvant versus adjuvant treatment.
While such immune-related adverse event could hamper on-time surgery, we have
not observed this within our OpACIN phase 1 trial, as all neo-adjuvant patients
have undergone lymph node dissection on time at week 6.
On the benefit side, of the 10 patients treated in the neo-adjuvant arm of
OpACIN, 5/10 patients had a PR, 2/10 SD, and 1/10 PD on CT-scan according to
RECIST 1.1after the first 2 courses (in 2 patients there was no CT-scan
performed before surgery). Moreover, 7/9 patients in the neoadjuvant arm
achieved a pathologic response, and none of the responders has relapsed. This
results currently in a 80% RFS rate and 90% OS rate at 25 months median
follow-up (minimum follow-up 15 months).
These observations have led to the design of this phase 2 trial that will test
only 2 courses of neo-adjuvant ipilimumab + nivolumab with tighter
discontinuation rules (Chapters 6.7 - 6.9), and in 3 different combination
schemes with the aim to reduce toxicity while preserving efficacy.
The design of the additional combination schemes to be tested (arm B and C) in
this study are based on several observations from other early phase trials in
stage 4 melanoma: a) in the phase 1 study testing ipilimumab + nivolumab in
late stage melanoma also 4 courses of a lower dose of ipilimumab (1 mg/kg) were
tested in combination with nivolumab 3 mg/kg indicating equally efficacy
(CA209-004) [11], b) data from the Keynote 029 study testing low dose
ipilimumab (1 mg/kg) with the standard dose of pembrolizumab (2 mg/kg) in 153
patients indicate that frequency of grade 3/4 toxicity is lower (to 42%), while
disease control is preserved (G. Long, ASCO 2016 meeting), and c) data from a
retrospective analysis of late stage melanoma patients treated at the NKI with
2 courses ipilimumab 3 mg/kg q3wks, directly followed the next day by
pembrolizumab or nivolumab at standard dosing (2 mg/kg q3wks or 3 mg/kg q2wks
respectively). This sequential scheme also induced a similar disease control
rate, as compared to the phase 3 trial testing the standard scheme ipilimumab +
nivolumab [12], but was less toxic. We observed in only 38% of the patients
grade 3/4 treatment-related toxicity [13]. Subsequently, we regularly treated
stage 4 melanoma patients with ipilimumab 3mg/kg for 1 or 2 courses, directly
followed by nivolumab or pembrolizumab, starting on the same day after the
ipilimumab infusion (2 hours observation interval), observing no additional
toxicities (manuscript in preparation).
Thus, the risk of neo-adjuvant immune related toxicity in OPACIN-neo was
considered to be lower (mainly in arm B or C) than that of patients within the
phase I OpACIN trial. In the planned interim analysis after 39 patients of
OpACIN-neo this was indeed the case observing 38%, 16%, and 46% grade 3/4
toxicities, respectively for the three arms. Unplanned additional interim
analyses in April and May 2018 (performed upon request of the DSMB and due to
report of a severe colitis in Arm C), reveled similar toxicity rates with 40%,
25%, and 50% for arm A, B, and C respectively. Arm C was closed earlier after
26 patients had been included upon the advice of the DSMB.
In addition, in OpACIN-neo patients are not allowed to be treated with local
radiotherapy, which bears the risk of reduced local tumor control. Considering
the fact that local irradiation only improves local tumor control, but does not
improve RFS or OS [1], participation in this trial might offer the chance for
improved recurrence free survival. This is based on the observation that
adjuvant ipilimumab has already shown RFS benefit in a prospective randomized
controlled phase 3 trial in stage III patients, the combination of ipilimumab
with nivolumab induces superior response rates as compared to ipilimumab alone
in stage IV melanoma [7, 11, 12, 14], and the high response rates observed in
the neo-adjuvant arm of the OpACIN trial.
In the PRADO extension cohort, patients will undergo resection of the marked
lymph node and pathologic response will be assessed. Only patients responding
on immunotherapy will not receive adjuvant radiotherapy. Non-responders may
receive adjuvant radiotherapy and adjuvant systemic therapy (nivolumab or
dabrafenib + trametinib). Patients achieving pCR or pnCR will not undergo
subsequent CLND. Patients without CLND might bear a higher risk of relapse,
which is the scope of this part of the trial to show that it is not the case.
At the same time, they can benefit from reduced surgery-related adverse event.
An additional burden as compared to standard therapy is, that the patients
possibly need to undergo an additional anesthesia if the location of the marked
lymph node is too deep for removal under local anesthesia.
To ensure the safety of these patients, a strict safety rule will be
implement