Recombinant MAGE-A3 Cancer Immunotherapy Promising

ATLANTA—A recombinant fusion protein designed to stimulate immune response against the MAGE-A3 tumor antigen has shown some efficacy in patients with completely resected early-stage non-small-cell lung cancer (NSCLC), decreasing the recurrence rate by about one-third. This improvement was seen mainly in patients with stage II disease, Johan Vansteenkiste, MD, PhD, of University Hospital Gasthuisberg, Leuven, Belgium, reported at the American Society of Clinical Oncology 42nd Annual Meeting (abstract 7019).

Dr. Vansteenkiste said that about 35% of stage IB-II NSCLC patients have tumors that express MAGE-A3. "It is tumor specific and is not expressed on normal cells," he said. Earlier work had suggested that MAGE-A3 expression might be associated with worse prognosis in NSCLC. MAGE-A3 is also expressed in 49% of head and neck cancers, 35% of bladder cancers, and 74% of melanomas. "Postoperative MAGE-A3 immunization might be a targeted, well tolerated, and effective treatment," he said.

This trial, sponsored by GlaxoSmithKline, combined MAGE-A3 as a purified, recombinant fusion protein with a GlaxoSmithKline proprietary AS02B adjuvant system. Patients with stage IB-II NSCLC who were MAGE-A3 positive by reverse transcriptase polymerase chain reaction (RT-PCR) underwent complete tumor resection, recovered from surgery to performance status 0 or 1, and were randomized to MAGE-A3 or placebo.

MAGE-A3 was given as 300 µg intramuscular injections. Induction included five injections every 3 weeks for 5 weeks followed by a maintenance phase of eight injections every 3 months. Total treatment time was 27 months. Placebo injections were given on the same schedule. The primary endpoint was time to recurrence. The planned interim analysis was carried out, according to protocol, a minimum of 18 months after enrollment of the last patients and after there were at least 28 recurrences. The trial enrolled 182 patients; 122 were randomized to MAGE-A3 and 60 to placebo.

MAGE-A3 treatment was well tolerated; the most common adverse effects were mild grade 1-2 local or systemic reactions lasting less than 1 day. There were 29 grade 3-4 events in 21 patients. The investigators concluded that three grade 3 adverse events were possibly related to treatment, and this led to withdrawal of two patients (one for local pain, one for an exacerbation of chronic obstructive pulmonary disease).

Dr. Vansteenkiste presented the preplanned interim analysis, with a median follow-up of 21 months. The recurrence rate was 30.3% with MAGE-A3 vs 41.7% with placebo (P = .138). When analyzed by stage, the recurrence rate in stage IB disease was 30.5% vs 33.3% (P = .835) and in stage II disease, 30.0% vs 57.1% (P = .055). "Subanalysis by stage should be interpreted with much caution, because of the limited number of patients, but the difference was about 27% in the stage II patients," he commented.

He also said that the data support concerns that sampling as a lymph node procedure may not be an adequate guide to treatment. In this study, lymph node sampling was done in 64 patients, and lymph node dissection in 118. The recurrence rates with sampling vs lymph node dissection in stage IB patients were 32.1% vs 29.6% with MAGE-A3 and 54.5% vs 25% with placebo, and in stage II patients, 46.7% vs 20.0% with MAGE-A3, and 70% vs 45.5% with placebo.

Cox regression analysis showed hazard ratios (HR) associated with MAGE-3 of 0.67 for all patients, 0.82 for stage IB patients, and 0.60 for stage II patients. None of these reached statistical significance. "The 33% reduction in relapse rate, HR of 0.67, for all patients is a positive signal of efficacy," Dr. Vansteenkiste said. "It suggests that this targeted approach is of potential benefit in NSCLC." He recommended testing the vaccine in combination with postoperative chemotherapy or as an alternative for patients unable to tolerate postoperative chemotherapy.