ESMO 2014: Adjuvant Treatment with MAGE-A3 Cancer Immunotherapeutic in Patients with Resected NSCLC Does Not Increase Disease-Free Survival

Results of the MAGRIT, a double-blind, randomised, placebo-controlled phase III study

The MAGRIT global trial assessed the efficacy of the recMAGE-A3 + AS15 cancer immunotherapeutic as adjuvant therapy in patients with resected MAGE-A3-positive non-small cell lung cancer (NSCLC). In this study, treatment with MAGE-A3 cancer immunotherapeutic did not increase disease-free survival (DFS) compared to placebo in either the overall population or in patients who did not receive adjuvant chemotherapy. The results were presented by Prof. Johann Vansteenkiste of the Respiratory Oncology Unit, University Hospitals Leuven - Campus Gasthuisberg, Leuven, Belgium during the Presidential Symposium 1 at ESMO Congress 2014 in Madrid Spain.

Adjuvant chemotherapy is the standard of care for stage II and IIIA NSCLC, and for high risk stage IB NSCLC. However, the 5-year DFS rate remains poor (35-50%) and about half of the patients will not receive adjuvant chemotherapy for various reasons. Tolerability of adjuvant cisplatin-based chemotherapy is suboptimal.

MAGE-A3 is a tumour-specific antigen. It is expressed in several tumour types, including NSCLC. MAGE-A3 cancer immunotherapeutic is delivered as a recombinant protein, combined with immunostimulants.

MAGE-A3 cancer immunotherapeutic showed activity in metastatic melanoma. Double-blind, placebo-controlled, phase II trial in completely resected MAGE-A3-positive stage IB-II NSCLC showed at 44 months follow-up, a 25% reduction in the relative risk of cancer recurrence with hazard ratio (HR) 0.75. It was very well tolerated. Predictive gene signature was discovered in metastatic melanoma and reproduced in NSCLC.

The MAGRIT findings

MAGRIT was a randomised, double-blind, placebo-controlled trial that investigated whether the recMAGE-A3+AS15 cancer immunotherapeutic as adjuvant therapy improves DFS in patients with completely resected (R0), anatomically resected, MAGE-A3-positive NSCLC (stages IB, II, and IIIA according to TNM classification version 6) who did or did not receive adjuvant chemotherapy (up to 4 cycles of platinum-based regimen), ECOG performance status (PS) 0, 1 or 2, adequate bone-marrow reserve, renal function and hepatic function and no auto-immune disease.

MAGE-A3 status was assessed in primary tumour by RT-PCR test on formalin-fixed paraffin embedded tissue.

Patients were randomly assigned (2:1) to receive 13 intramuscular injections of MAGE-A3 cancer immunotherapeutic or placebo over a 27-month treatment period.

The three co-primary endpoints were DFS in the overall and in the no-adjuvant chemotherapy population and DFS in patients with a potentially predictive gene signature.

Secondary endpoints included overall survival (OS), lung cancer specific survival, disease-free specific survival, immunogenicity, safety, and health-related quality of life.

In total 13,849 patients were screened and 4,210 of them had a MAGE-A3 positive tumour sample. However, 2,272 patients were randomised and treated.

One interim analysis was done which concluded that the study may continue as pre-specified boundary was not met and the treatment was well tolerated.

Overall, 52% of the patients received adjuvant chemotherapy. Stage IB disease was recorded in 47%, stage II in 36% and 17% were stage IIIA, respectively. Median age was 63 years and 24% of patients were females.

Mean relative dose intensity was above 98% in both groups throughout the treatment period.

Common adverse events present in more than 10% of patients treated with MAGE-A3 cancer immunotherapeutic vs placebo were:

  • pyrexia (35% vs 5%),
  • injection site pain (31% vs 5%),
  • injection site reaction (18% vs 14%),
  • fatigue (16% vs 7%),
  • pain (16% vs 2%),
  • influenza-like ilness (13% vs 3%),
  • and myalgia (12% vs 2%).

However, the rate of grade ≥ 3 adverse events did not differ between treatment groups and was below 1%.

Median follow-up at the time of final analysis was 38.8 months. In the overall study population, median DFS was 60.5 months and 57.9 months respectively for MAGE-A3 cancer immunotherapeutic and placebo (HR 1.024, p = 0.7379). In patients who did not receive adjuvant chemotherapy, median DFS was 58.0 months and 56.9 months for MAGE-A3 cancer immunotherapeutic and placebo groups, respectively (HR 0.970, p = 0.7572).

Overall survival in the overall popuation was not reached, but it might be expected to exceed median value of 5 years.

Due to the absence of treatment effect, a gene signature predictive of clinical benefit to MAGE-A3 cancer immunotherapeutic could not be identified.


The authors concluded that MAGRIT is the largest clinical trial in NSCLC and the first one to investigate immunotherapy in the adjuvant setting of early stage NSCLC. Adjuvant treatment with the MAGE-A3 cancer immunotherapeutic did not increase DFS compared to placebo in the overall population nor in patients who did not receive adjuvant chemotherapy. No benefit was observed in any subset analyses. MAGE-A3 cancer immunotherapeutic was generally well tolerated with mainly mild toxicities and no detectable increase in immune-mediated disorders. No predictive gene signature was identified in the training set. The study database is a source for further analysis on global contemporary approach to early stage NSCLC.

Dr George Coukos, who discussed the study results, said that the advantages of vaccine therapy in NSCLC are low toxicity, easiness of administration, and ability to induce anti-tumour protective memory. The disadvantages are low impact at present time. Clinical indication might be in the maintenance/consolidation setting.

The normal function of MAGE-A3 is unknown, but its presence on tumour cells has been associated with a worse prognosis. The MAGE-A3 antigen is expressed in a variety of tumour cells, but not in normal tissues (except for the testis). In NSCLC, expression can be demonstrated in 35% of early-stage tumours. The ‘MAGE-A3 vaccine’ is an example of a recombinant protein antigen-based vaccine.

Possible reasons for study failure, according to Dr Coukos, are retrospective subset analysis that may be deceiving, monovalent molecularly defined vaccines might be too weak to make an impact, metastatic or progressive tumours may be immune escape variants, and/or it could be due to epigenetic mechanisms (association between methylation  and lung cancer recurrence).

Dr Coukos discussed future of vaccines in NSCLC, in particular awaiting the results from other monovalent molecularly defined vaccines and developing polyvalent molecularly defined vaccines.

High mutational rates may contribute to increased immunogenicity. Melanomas and lung tumours display many more mutations than average, with approximately 200 non-synonymous mutations per tumour. These larger numbers reflect the involvement of potent mutagens. Lung cancers from smokers have 10 times as many somatic mutations as those from non-smokers.

Therefore, the future opportunities for vaccines in NSCLC, according to Dr Coukos, are personalised molecular vaccines based on mutanome analysis and autologous whole tumour antigen vaccines designed to address specific mutations.


GlaxoSmithKline Biologicals SA was the funding source in all stages of the study/project conduct and analysis.