A Phase III, Open-label, Randomised, Multi-centre, International Study of MEDI4736, Given as Monotherapy or in Combination with Tremelimumab, Determined by PD-L1 Expression, Versus Standard of Care in Patients with Locally Advanced or Metastatic Non-Small Cell Lung Cancer (Stage IIIB IV) who Have Received at Least Two Prior Systemic Treatment Regimens Including One Platinum based Chemotherapy Regimen and Do Not Have Known EGFR TK Activating Mutations or ALK Rearrangements (ARCTIC)

Lung cancer has been the most common cancer in the world for several decades,
and by 2008, there were an estimated 1.61 million new cases, representing 12.7%
of all new cancers. It was also the most common cause of death from cancer,
with 1.38 million deaths (18.2% of the total) (GLOBOCAN 2008). Non-small cell
lung cancer (NSCLC) represents approximately 80% to 85% of all lung cancers.
Unfortunately, at the time of diagnosis, approximately 70% of patients with
NSCLC already have advanced or metastatic disease not amenable to surgical
resection. Furthermore, a significant percentage of patients with early stage
NSCLC who have undergone surgery subsequently develop distant recurrence and
die as a result of their lung cancer (Pisters and Le Chevalier 2005).

Despite advances in the diagnosis, imaging, staging and treatment of NSCLC, the
estimated overall 5-year survival for patients in Europe continues to be low
(11%) (D*Addario et al 2010). Once patients have treatment failure following
initial therapy, the outlook for those with refractory advanced NSCLC is
extremely poor, with response to further systemic treatment of <10% (GLOBOCAN
2008, Hanna et al 2004) and median survival of approximately 6 months.

Common third-line treatment for NSCLC in major global markets includes:
vinorelbine (NAVELBINE®), tyrosine kinase inhibitors (TKIs such as erlotinib
[TARCEVA®] and gefitinib [IRESSA®]), pemetrexed (ALIMTA®), and docetaxel
(TAXOTERE®) for non-squamous NSCLC, and vinorelbine, TKIs, gemcitabine (GEMZAR)
and docetaxel (TAXOTERE®) for squamous NSCLC (Decisions Resources 2013). For
these patients, clinical trials, experimental treatment, or best supportive
care are among the treatment options (Azzoli et al 2009, Syrigos et al 2011).

The immune system can identify tumour-associated antigens and eliminate the
cancerous cells expressing them and thus plays an important role in preventing
and combating the growth of tumours. Blockade of negative regulatory signals to
T-cells such as cytotoxic T-lymphocyte antigen 4
(CTLA-4) and programmed death ligand 1 (PD-L1) has shown promising clinical
activity. PD-L1 is expressed in a broad range of cancers with a high frequency,
up to 88% in some types of cancer.

In vitro, an antibody that blocks the interaction between PD-L1 and its
receptors can relieve PD-L1-dependent immunosuppressive effects and enhance the
cytotoxic activity of anti-tumour T-cells (Blank et al 2006). Based on these
findings, an anti-PD-L1 antibody could be used therapeutically to enhance
anti-tumour immune responses in patients with cancer. Results of several
preclinical studies using mouse tumour models support this hypothesis, where
antibodies directed against PD-L1, or its receptor PD-1, showed anti-tumour
activity (Hirano et al 2005, Iwai et al 2002, Okudaira et al 2009, Zhang et al
2008).

For more details please also refer to chapter 1. INTRODUCTION, of Clinical
Study Protocol D4191C00004 Version 2 29 August 2014.

This study is a Phase III, randomised, open label, multi-centre study assessing
the efficacy and safety of MEDI4736 versus Standard of Care in NSCLC patients
with PD-L1-positive tumours and the combination of MEDI4736 plus tremelimumab
(MEDI4736+tremelimumab). The study will enrol male and female patients with
locally advanced or metastatic NSCLC (Stage IIIB-IV), who have received at
least 2 prior systemic treatment regimens including 1 platinum-based
chemotherapy regimen for NSCLC. Patients with known epidermal growth factor
receptor (EGFR) tyrosine kinase (TK) activating mutations and anaplastic
lymphoma kinase (ALK) rearrangements are not eligible for the study
(prospective testing is not planned within this study).

This study will consist of 2 sub-studies: Sub-study A will address the research
hypotheses for MEDI4736 monotherapy in patients with programmed death ligand 1
(PD-L1)-positive tumours and Sub-study B will address the research hypotheses
for MEDI4736 plus tremelimumab (MEDI4736+tremelimumab) in patients with
PD-L1-negative tumours.

Patients will be randomised in a 1:1 ratio in Sub-study A and a 1:1:1:1 ratio
in Sub-study B.

Sub-study A (patients with PD-L1-positive tumours)
* MEDI4736 (10 mg/kg Q2W iv for up to 12 months) (125 patients)
* Standard of care (restricted to the erlotinib, gemcitabine or vinorelbine)
(125 patients). For each agent 4 weeks equates to 1 cycle of treatment.
* Erlotinib: 150 mg once daily as a tablet for oral administration
taken at least 1 hour before or 2 hours after the ingestion of food
* Gemcitabine: 1000 mg/m2 iv over 30 minutes on Days 1, 8, and 15 of a
28-day cycle
* Vinorelbine: 30 mg/m2 iv on Days 1, 8, 15 and 22 of a 28-day cycle.

Sub-study B (patients with PD-L1-negative tumours)
* MEDI4736+tremelimumab (MEDI4736 20 mg/kg plus tremelimumab 1 mg/kg once
every 4 weeks [Q4W] iv for up to 12 weeks [4 doses]) then MEDI4736 alone (10
mg/kg Q2W iv, starting at Week 16, for 34 weeks [18 doses]) (180 patients)
* Standard of care (see sub-study A) (120 patients)
* MEDI4736 (10 mg/kg Q2W iv for up to 12 months) (120 patients)
* tremelimumab (10 mg/kg Q4W iv for 24 weeks then Q12W for a further 24 weeks)
(60 patients)

The sub-studies may not run concurrently with completion of recruitment
potentially occurring at different time points. Assignment to the applicable
sub-study will be preceded by the Pre-screening Period during which assessment
of the patient*s PD-L1 status, based on a tumour sample, will take place. After
confirmation of PD-L1 status, patients will enter the main Screening Period
within their assigned sub-study if it remains open for recruitment.

The primary objective of this study is to assess the efficacy of MEDI4736
monotherapy and MEDI4736+tremelimumab compared with Standard of Care in terms
of overall survival (OS) and progression free survival (PFS).

Please refer to the PROTOCOL SYNOPSIS for more details on study design.

Sub-study A (patients with PD-L1-positive tumours)
* MEDI4736 (10 mg/kg Q2W iv for up to 12 months)
* Standard of care (restricted to the erlotinib, gemcitabine or vinorelbine).
For each agent 4 weeks equates to 1 cycle of treatment.
* Erlotinib: 150 mg once daily as a tablet for oral administration
taken at least 1 hour before or 2 hours after the ingestion of food
* Gemcitabine: 1000 mg/m2 iv over 30 minutes on Days 1, 8, and 15 of a
28-day cycle
* Vinorelbine: 30 mg/m2 iv on Days 1, 8, 15 and 22 of a 28-day cycle.

Sub-study B (patients with PD-L1-negative tumours)
* MEDI4736+tremelimumab (MEDI4736 20 mg/kg plus tremelimumab 1 mg/kg once
every 4 weeks [Q4W] iv for up to 12 weeks [4 doses]) then MEDI4736 alone (10
mg/kg Q2W iv, starting at Week 16, for 34 weeks [18 doses])
* Standard of care (see sub-study A)
* MEDI4736 (10 mg/kg Q2W iv for up to 12 months)
* tremelimumab (10 mg/kg Q4W iv for 24 weeks then Q12W for a further 24 weeks)

Agents that act via antagonism of an inhibitory pathway modulate an existing
antigen-specific T-cell receptor signal and have a limited potential to drive
systemic, nonspecific activation of T cells. MEDI4736 antagonizes an inhibitory
receptor (PD-L1) and as such, in the absence of an antigen-specific T-cell
receptor signal, inhibition of function of PD-L1 is not anticipated to elicit
any response. MEDI4736 did not induce release of any cytokine from any donor at
any concentration tested. Experience with MEDI4736 is limited, but for the 20
patients treated to date with available safety data (in the dose-escalation
phase of the study on a Q2W schedule), there have been no DLTs. The majority of
AEs (in 15 of the 20 patients) have been CTCAE Grade 1 or Grade 2. There have
been no Grade 3 or higher treatment-related AEs. Six patients have had a total
of 11 treatment-emergent SAEs. Four patients have died due to AEs but none of
these events were considered by the reporting investigator to be related to
treatment with MEDI4736.

The potential for clinical benefit associated with inhibition of the PD-1/PD-L1
pathway, supported by objective responses observed in earlier studies in
patients with NSCLC, outweighs the known and potential risks based on the AEs
reported in patients treated with MEDI4736 and other PD-1/PD-L1 inhibitors.
Thus, the benefit/risk assessment, favours the conduct of this proposed study.