ESMO @ ECC 2013: A Phase III Study of Tivantinib Plus Erlotinib Did Not Meet a Primary Endpoint in Patients with Locally-advanced or Metastatic, Non-squamous NSCLC

Analysis of molecular subgroups, including MET expression, is ongoing.

Date: 29 Sep 2013

Although a large, phase III trial of tivantinib plus erlotinib for the treatment of patients with locally-advanced or metastatic, non-squamous, non-small cell lung cancer (NSCLC) did not meet its primary endpoint of prolonging overall survival, promising progression-free survival and overall response rate results were seen in these patients. Following trial cessation by an independent review board, an analysis of data by molecular subgroup is underway. From the molecular analysis, it is expected to answer whether clinical benefit in patients with tumours that overexpress MET is observed.

Findings were reported 28 September, 2013 by Dr. Giorgio Scagliotti of the Oncology Department, University of Turin San Luigi Hospital, Orbassano (Turin), Italy during the Lung Cancer Proffered Papers Session (Abstract E17-1821) at the 17th ECCO – 38th ESMO – 32nd ESTRO European Cancer Congress. The European Cancer Congresses are organised in joint partnership with ESSO, EACR, EONS and SIOPE as a forum for the latest in oncology research and to offer multidisciplinary and multi-professional educational opportunities in oncology. The 2013 Congress convened in Amsterdam, The Netherlands from 27 September until 1 October.

A phase III MARQUEE trial explored the efficacy of tivantinib (selective MET inhibitor) and erlotinib (EGFR inhibitor) vs. placebo and erlotinib in previously treated non-small cell lung cancer patients with tumours of non-squamous histologies. The study results were awaited by the oncology community with particular attention due to promising data of this combination in previously reported phase II trial; in addition several preclinical models strongly suggested a possible clinical benefit. The trial was aimed to document improvement in overall survival with secondary endpoints of progression-free survival, efficacy according to molecular characteristics, and safety.
With 1048 patients recruited in just 19 months, the trial failed to meet its primary objective. The non-significant hazard ratio for overall survival was 0.98. Toxicity of combined tivantinib and erlotinib was similar to toxicity of placebo and erlotinib with the exception of higher rate of grade 3 or 4 neutropenia (10% vs. 1%, respectively). During the ECC 2013, Prof. Scagliotti presented important efficacy data in subsets of patients according to MET protein expression in tumour samples, with available and valid measurements in approximately 40% of trial participants. According to these subset analyses, both progression-free and overall survival was longer in patients treated with combined MET-EGFR inhibition when tumours showed at least 2+ positive MET immunostaining in more than 50% of tumour cells.
Inhibition of MET, a receptor for hepatocyte growth factor, is a promising therapeutic strategy in several malignancies, lung and gastric cancers in particular. Several phase I–III clinical trials testing MET inhibition strategies are ongoing in these and other cancer types. Undoubtedly, the MARQUEE adds important data to the question of optimal use of MET inhibitors in the future. Based on unfilled promise of this trial, we can no longer ignore the biological measures to identify patients most likely to benefit from MET inhibitor therapy. Although it is unclear which of the several biomarkers are most appropriate to select patients for combined MET and EGFR inhibition therapy, early phase data for tivantinib and other MET inhibitors clearly indicated a possible role of MET protein expression, MET gene copy number and KRAS mutations. The question of which biomarker to use prospectively, and how, should have been resolved before MARQEE was initiated.

-
Rafal Dziadziuszko*

*Rafal Dziadziuszko, ESMO spokeperson who was not involved in the study

Promising results seen in a randomised phase II study

Dr. Scagliotti headed an international team of investigators in conducting MARQUEE, a randomized, double-blind, placebo-controlled, phase III trial that evaluated the benefit of adding the c-MET inhibitor tivantinib (ARQ 197) to erlotinib compared to placebo plus erlotinib in previously treated patients with locally-advanced or metastatic, non-squamous NSCLC.

The trial was based upon promising results from a randomised phase II study that showed tivantinib plus erlotinib improved progression-free survival (PFS) and overall survival (OS) over placebo and erlotinib in a subset of patients with non-squamous histology, which is a population that is enriched for MET overexpression.

The MARQUEE study results

Beginning in January, 2011 the trial enrolled 1048 patients with locally-advanced or metastatic non-squamous NSCLC who had previously received one or two lines of systemic therapy, including a platinum-doublet; patients were excluded if received prior treatment with erlotinib or any other EGFR inhibitor.

Patient median age was 62 (range 24-89) years; 66% and 34% of patients, respectively, had received one or two prior therapies and the ECOG performance status was 0 in 32% or 1 in 68% of patients. EGFR mutations were confirmed in 10.4% of the patients and 27.1% of patients harboured KRAS mutations. These characteristics were well balanced between treatment arms.

The patients were stratified by the number of prior therapies, sex, smoking history and EGFR and KRAS mutation status and then randomized 1:1. Oral tivantinib was given at 360 mg twice daily plus erlotinib at 150 mg once daily to 526 patients while 522 patients received the same dose of erlotinib plus placebo until disease progression.

The primary endpoint was OS; an interim analysis was preplanned that included limits for futility or superiority, which were intended to end the trial or allow cross-over. Secondary and exploratory endpoints included PFS, OS in molecular subgroups and safety.

Results from the pre-planned interim analysis did not show a significant patient benefit with adjunct tivantinib and the independent data monitoring committee recommended trial discontinuation in September, 2012 because the data had crossed the futility boundary. At the data cut-off in December, 2012 the median OS was 8.5 months with the tivantinib/erlotinib combination compared to 7.8 months with placebo plus erlotinib (hazard ratio [HR] 0.98; p = 0.81).

However, a trend toward clinical benefit was demonstrated in the tivantinib/erlotinib arm in the PFS and overall response rate (ORR) data; median PFS significantly favoured adjunct tivantinib and was 3.6 months with tivantinib/erlotinib compared to 1.9 months with placebo/erlotinib (HR 0.74; p < 0.0001). The ORR was also improved in the tivantinib/erlotinib arm with 10.3% of patients demonstrating a response compared to 6.5% of patients receiving placebo/erlotinib (p < 0.05).

The tivantinib/erlotinib combination was well tolerated and showed a safety profile similar to placebo/erlotinib. Adverse events (AEs) included rash which was commonly reported by 33.1% of tivantinib/erlotinib versus 37.3% of placebo/erlotinib patients. Other AEs were diarrhoea, reported by 34.6% versus 41.0% and asthenia/fatigue by 43.5% versus 38.1% of tivantinib/erlotinib and placebo/erlotinib patients, respectively. Only grade 3/4 neutropenia was more common with tivantinib and occurred in 10.0% tivantinib/erlotinib versus 1.0% of placebo/erlotinib patients.

The investigators concluded that, while the trial did not reach the primary endpoint of improved OS with the addition of tivantinib to erlotinib, the PFS and ORR data did show that these patients experienced substantial improvement. This promising biological activity plus the tolerability of the combination treatment encouraged them to perform the preplanned analysis of OS, PFS and ORR by molecular subgroups, including MET expression.

The MARQUEE trial was supported by ArQule Inc., Daiichi Sankyo and the Experimental Cancer Medicine Centre.

Last update: 29 Sep 2013

No conflicts of interest for the analysis were reported.