ESMO 2014 - Abstract LBA2_PR

ESMO 2014 Abstract LBA2_PR: Gefitinib/chemotherapy vs chemotherapy in epidermal growth factor receptor (EGFR) mutation-positive non-small-cell lung cancer (NSCLC) after progression on first-line gefitinib: the Phase III, randomised IMPRESS study.

Tony Mok et al. (Hong Kong China)

Buenos días de Madrid,

 On Sunday October 28th, the first of two Presidential Symposiums highlighted two important randomized phase III trials conducted in Non-Small Cell Lung Cancer : the adjuvant recMAGE-A3 vaccine trial MAGRIT and the IMPRESS (Iressa Mutation Positive Multicentre Treatment Beyond Progression Study) trial.

 In short, the 2272  patient MAGRIT trial is one of the largest randomized trial conducted in NSCLC but failed to show any clinical impact in terms of disease progression and overall patient survival. 

 The last presentation of the day was thereafter delivered by Dr. Tony Mok. The rationale for IMPRESS was to determine a preferable strategy of treatment in patients with EGFR mutated NSCLC after failure from first line treatment with gefitinib. As the use of second-line chemotherapy is considered the standard of care for this situation, the investigators hypothesized that the continuation of gefitinib in addition to the chemotherapy backbone may improve outcomes. This study was performed in 71 centres in Europe and Asia and randomized 265 patients in a 1 :1 fashion to cisplatin+pemetrexed compared to cisplatin+pemetrexed+gefitinib at standard doses. Of note, patient selection was based on demonstrated sensitivity to first-line gefitinib. Only those achieving CR/PR > 4 months or SD > 6 months with gefitinib were deemed eligible. The primary endpoint was progression-free survival and was powered to detect a PFS HR < 0.63.

 The presenter highlighted slight imbalances in the patient distribution disadvantaging the experimental arm. Age greater or equal to 65 years and the presence of brain metastases at baseline was found in 32% and 33% in the triple therapy arm and 26% and 23% in the chemotherapy doublet arm, respectively.

 The tumor response rate was 31.6% in the gefitinib arm and 34.1% in the placebo arm (n.s.). The PFS curves showed that the gefinitib arm was barely on top of the placebo arm, but at the median, the duration of PFS was identical at 5.4 months (HR=0.86, p=0.273). The Forest plot of the subgroups of interest of this PFS analysis did not demonstrate any particular group that appeared to fare better with gefitinib. Overall survival was presented with only 33% of events reached so far. This preliminary OS analysis suggests a possible detrimental effect of gefitinib (HR=1.62, p=0.029). Median OS was 17.2 months in the placebo arm and 14.8 months in the gefitinib arm. The OS results were also presented after an ad-hoc analysis was conducted to include the presence of brain metastases as a covariate and still showed suggestions of a detrimental effect of gefitinib (HR=1.55, p=0.05). More patients on the placebo arm were rechallenged to an EGFR-TKI after study therapy discontinuation (33% vs 23%). Toxicity evaluation showed more rash and GI toxicities in the gefitinib arm.


CARE Faculty Perspective: The use of chemotherapy has not been formally shown to improve overall survival in the post-TKI setting but has been considered a standard by experts. Considering the biology of this disease, many experts believed that chemotherapy may be effective against resistant clones and that continuation of the TKI during chemotherapy may prevent the reappearance of the EGFR-sensitive clones, leading to greater benefit than with chemotherapy alone. In this regard, the results of the IMPRESS study seem counter-intuitive and surprising. Although PFS was not negatively impacted by the use of chemotherapy with gefitinib, the reasons why OS may be impaired by this combination was debated after the presentation of IMPRESS. Potential explanations may include the immaturity of the OS results, the imbalances in the treatment arms as well as toxicity of the triple drug combination, perhaps leading to increased decline in performance status and/or less access to third line treatment for this group. Of note, the phase II LUX-LUNG 5 studied paclitaxel vs paclitaxel+afatinib in this setting and did not show differences in OS between both arms.

In the end, the IMPRESS trial does show that chemotherapy alone can offer to a 34.1% response rate and a median OS of 17.2 months in a second line setting and that combination of EGFR-TKIs with chemotherapy should be avoided considering the current evidence.

Normand Blais, MD