MTE08.01 Immunotherapy in Early and Locally Advanced NSCLC: Challenges and Perspectives

Abstract

The demonstration that therapies directed at the programmed death‐1 (PD‐1) receptor or its ligand (PD‐L1) result in durable responses and improved survival in a number of solid tumors including non‐small cell lung cancer has awakened interest in cancer immunotherapy. The activity of PD‐1/PD‐L1 therapy in NSCLC implies that endogenous T‐cells can recognize antigens on tumor cells and eliminate those cancer cells. The success of checkpoint inhibitor therapy in the metastatic setting has led to a immunotherapy trials in early stage (adjuvant) and stage lll NSCLC. This session will provide perspective on the current state and challenges facing immunotherapy in these settings. Current perspectives: The concept of using immunotherapy to prevent recurrence of NSCLC after resection of early stage NSCLC is not new. More recently, two randomized phase lll trials of therapeutic cancer vaccine strategies have been completed in resected, early stage NSCLC (MAGRIT) or after chemoradiation in stage lll disease (START). The MAGRIT trial1 assessed the efficacy of an active, specific cancer immunotherapy (ASCI) against the MAGE‐A3 cancer testis antigen in completely resected stage IB‐IIIA NSCLC. Tumors from more than 13,000 patients were screened for MAGE‐A3 expression and 2312 patients whose tumors expressed MAGE A3 were randomized 2:1 to MAGE‐A3 (ASCI) or placebo. The MAGRIT trial failed to meet its primary end‐point of improvement in disease free survival with MAGE‐A3 ASCI. The START trial2 assessed a MUC1 vaccine in stage III NSCLC patients who had response or stable disease after standard chemoradiation. The chemoradiation could have been delivered concurrently or sequentially. The modified intention to treat population included 1239 patients. The primary end‐point was not met (adj. HRO .88, 95% CI 0.75 ‐103, p=0.123). Further development of this agent has been abandoned. The failure of these two large global phase III studies raises doubt about vaccine strategies used in isolation in early stage NSCLC. There are a number of possible explanations for these negative results.3 One of the primary reasons is that cancer vaccines, when used alone, fail to address the many immunosuppressive factors operating in the tumor microenvironment (TME). Clinical trials evaluating anti PD‐1/PD‐L1 therapy in early stage or locally advanced NSCLC have not yet reported results. The PACIFIC trial (NCT 20125461) is a randomized phase III trial of MEDI4736 versus placebo following concurrent chemoradiation in patients with stage III NSCLC. The primary outcome measures are OS and PFS. This trial completed accrual in April 2016 and has randomized more than 700 patients. In addition to the important efficacy outcomes, a number of exploratory objectives will assess tissue and blood for potential biomarkers. The Canadian Cancer Clinical Trials Group is assessing MEDI 4736 versus placebo in completely resected stage IB‐IIIA NSCLC (NCT 02273375). This trial will randomize 1100 patients with the primary outcome measure being DFS in PD‐L1 positive patients. PD‐L1 positive is defined as > % positive tumor cells. Immune based prognostic markers: The TME consists of stromal cells including endothelial cells and fibroblasts and a number of immune cell types. Tumors may escape immune recognition in large part by modulating the recruitment and function of various immune cells into the TME.4 A comprehensive review of the prognostic value of different immune cells in NSCLC has been reported.5 Two recent studies have separately assessed tumor lymphocytic infiltration (TLI)6 or stromal CD8+ T‐cell density7 as potential prognostic markers in early stage NSCLC. Using the large, relatively homogenous population of curative resected NSCLC patients from the LACE‐Bio collaboration, Brambilla et al examine the prognostic and predictive value of TLI. Patients were separated into discovery and validation sets. An intense TLI (>50% stromal lymphocytes in tumor bulk) was strongly prognostic of favorable overall survival and disease free survival. Based on previous work, Donner et al selected stromal CD8+ tumor infiltrating lymphocyte as the most promising immunobased prognostic marker. Using four separate cohorts of curatively resected stage I‐III patients, they established training and validation sets. Tissue microarrays were scored for stromal CD8 TLI's; stromal CD8 TIL density was found to be an independent prognostic factor and retained significant prognostic impact within each stage. The value of PD‐L1 as a biomarker in NSCLC has been investigated primarily in advanced disease and focused on prediction of response and/or survival. Studies investigating the value of PD‐L1 as a prognostic marker in early stage NSCLC have many limitations. These studies are small, include heterogenous populations, assess PD‐L1 using different antibodies and scoring systems and included PDL1 on tumor cells only or tumor cells and TLI's. It is not surprising that these studies show conflicting results. Based on the available evidence, the prognostic value of PD‐L1 expression in early stage NSCLC remains uncertain. The adjuvant trials of anti‐PD1/PD‐L1 therapy currently being conductedmay clarify the value of PD‐L1 as both prognostic and predictive biomarkers in this setting. Challenges One of the fundamental challenges to developing effective cancer immunotherapies is our limited understanding of the human immune system in steady state and its response to stress. Animal models do not necessarily translate to humans. The Human Vaccines Project8 is a global initiative that has as one of its primary objectives the decoding of the human immune system and providing a map of the human “immunome”. This private‐public partnership uses state‐of‐the‐art machine learning and technologies to elucidate the principles of immunogenicity to accelerate the development of new immunotherapies against infectious diseases and cancer. A second challenge is howbest to target micrometastases in the adjuvant and locally advanced setting. While the primary tumor and metastatic lesions have many mutations in common, metastatic tumors possess mutations that are distinct from the primary. Do adjuvant therapies need to target the metastatic cascade and if so, which steps are the most susceptible to intervention? 9 The complex of the TME would predict that focusing on TIL's or PD‐L1 is likely to result in only modest improvements in outcome. Blank et al10 argue that it will take a combination of biomarkers, the “cancer immunogram,” to determine the best approach in individual patients.