Current chemotherapy for advanced non-small cell lung cancer, not amenable for curative local treatment (surgery or chemoradiotherapy), has a modest life-prolonging effect and can improve quality of life. There is however no potential for long-term cure for these patients.

Chemotherapy also produces variable and often significant toxicity. Current retrospective evidence suggests that significant clinical responses can be obtained when patients whose cancer cells have an EGFR TKD mutation are treated with an EGFR TKI.

The ease of administration and toxicity profile of TKI compare favourably with that of chemotherapy, even single agents such as for example gemcitabine The present study will establish the clinical benefit rate of TKI as a first line treatment in patients with EGFR mutations and thus estimate the proportion of patients who might benefit for a prolonged period from a treatment with a modest toxicity profile.

Patients with stage IV NSCLC and some patients with advanced locoregional disease (stage III) are in general incurable and have a low probability for long-term survival.

Current systemic treatment in good PS patients (PS 0-1) consists of a cisplatin doublet (e.g. cisplatin plus a second drug: vinorelbine, gemcitabine, paclitaxel, docetaxel). The median and one year survival obtained with this treatment ranges between 8-10 months and 25 - 35 % respectively (1,2). Progression-free survival is a median of 5 month or less in randomized studies (3). Response rates obtained are less than 25% in metastatic disease and up to 75% in advanced locoregional disease.

Receptor tyrosine kinase inhibitors (RTKI's) are active drugs in patients with NSCLC pre-treated with cisplatin and/or docetaxel containing chemotherapy (2,4). Recently it has been shown that the EGFR1 kinase inhibitor erlotinib added to best supportive care (BSC) prolonged survival compared to BSC alone in patients with advanced NSCLC failing 1st or 2nd line chemotherapy (5).

A small number of preliminary reports have indicated that the objective response rate with these RTKI's as first line treatment in some patient populations with advanced NSCLC could be around 20 % (6), The objective (mainly partial) response in the phase II studies with pre-treated patients ranges from 10 - 15 % according to the level of pre-treatment but the control of the disease (stable disease) and improvement of symptoms without demonstrated objective response has been reported to be as high as 40 to 50 % (2,4). The addition of RTKI's or placebo to the current standard doublets (e.g. cisplatin, gemcitabine or taxol/carboplatin) has not been shown to impact on response rate, time to treatment failure or survival in large phase III randomised trials in advanced NSCLC patients not selected for EGFR expression (7,8).

Recently, mutations in the intracellular EGFR kinase domain that increase sensitivity of the receptor to RTKI's have been discovered (9,10). These studies suggest that the response/resistance to treatment could be strongly correlated to the presence/absence of such mutations. It is presently unclear whether patients who achieve stable disease under RTKI treatment do have receptor mutations, of which nature such mutations could be or what other biological pathways modulate the responsiveness/resistance.

Mutations have been observed in about 10 % or less of the tumours examined. Most mutations in the EGFR kinase domain have been found in adenocarcinoma of non-smokers or minimal smokers.

Recent data have established that an EGFR mutation can be found in 30% of adenocarcinoma of the lung if the smoking history is maximally 15 years. The probability of finding a mutation is approximately 50% in never-smokers with adenocarcinoma (11).

In contrast to the rather low objective response rate, the clinical benefit rate of treatment with RTKI is around 40 % in a general population of non-small lung cancer (9,10), more than can be accounted for by mutational analysis. EGFR gene amplification might be another determinant for sensitivity, but the biological factors leading to disease stabilisation in some patients with non-activated EGFR are presently unknown.

Several humanized monoclonal antibodies (huMoAb) that can inhibit the EGFR-1 receptor have been developed as well. Of these, cetuximab (ErbituxÒ) has been developed most extensively, especially in head and neck (12) and colorectal cancer (13,14), but other, fully humanized antibodies (huMoAb) (ABX-EGF, EMD 72000) are under development.

Cetuximab has been assessed in NSCLC, especially in combination with first-line and second-line chemotherapy (15), but no data are currently available with respect to single agent activity.

The two types of currently available drugs targeting the EGFR-1, the RTKI's such as gefitinib and erlotinib and huMoAb such as cetuximab and ABX-EGF, are generally well tolerated and devoid of significant grade 3-4 toxicity.

While there are strong indications that the major responses obtained with RTKI's could strongly correlate with the presence of tyrosine kinase domain mutations in addition to the known mechanism of action of these drugs (16,17), this is at present yet not so clear for the monoclonal antibodies directed against the extracellular domain of the receptor.

Some recent data suggest that EGFR gene amplification might also correlate with response to TKI (18).

Patients with Stage IV and some advanced stage III NSCLC are at present incurable and first-line chemotherapy has had only a palliative and a life prolonging effect at a cost of variable toxicity.

On the other hand the response rate to TKI in patients with a NSCLC carrying a mutant receptor is probably high as can be inferred from the comparison of the prevalence of such mutations and the response rate to these agents in an unselected population (16).

It seems therefore reasonable to explore in a first-line setting the potential of EGFR-1 targeted therapy, as these drugs have a very favourable toxicity profile and might induce prolonged palliation.

There are currently no data that would suggest that delaying chemotherapy in this patient population in well controlled conditions and for a limited time period (sufficient to evaluate the efficacy of more novel first-line treatments) might affect the efficacy of subsequent chemotherapy. There are also no data that suggest that prior administration of drugs targeted at the EGFR-1 will induce resistance to subsequent chemotherapy. The distinct respective mechanisms of action of these drugs and chemotherapy also do not suggest this to be likely.

In other diseases such as prostate cancer and oestrogen receptor-positive breast cancer, the availability of hormonal treatments permits prolonged and at times long-term disease control with limited discomfort. It is possible that a similar scenario could be obtained with an anti-EGFR treatment in NSCLC carrying a mutant EGFR.

Therefore it seems justified to expand previous second-line phase II experience with anti-EGFR strategies to patients with incurable stage III-IV NSCLC without prior therapy and a mutant EGFR. The informed consent towards these patients before entry in the study will clearly state that until now the first-line standard of care has been chemotherapy, which leads to a small overall survival benefit compared to the supportive treatment only and that participation in the current study will delay this treatment within a tightly controlled setting.

In the current trial the RTKI erlotinib will be used. Prior experience with this drug has been obtained in phase II and III studies. In the phase II setting, in an unselected pretreated population, a response rate of 12.3% and a median survival of 8.4 months could be obtained (19). In phase III, placebo-controlled studies, combination of erlotinib with concurrent chemotherapy does not provide a survival advantage (gemcitabine and cisplatin in TALENT (20) and paclitaxel and carboplatin in TRIBUTE (21). Several hypotheses can explain these results: lack of selection of patients; antagonism between cytostatic and cytotoxic agents (negative interaction with chemotherapy when given concurrently); chemotherapy and EGFR inhibitors target the same cell population since chemotherapy directly or indirectly affects EGFR function/expression and thereby reducing the effects of EGFR inhibitors. These results are in contrast to the significant survival benefit seen in the phase III randomized placebo-controlled study in NSCLC following failure of 1st or 2nd line chemotherapy (BR21). The response rate was 9% in the erlotinib treated patients, with significant increase in progression free survival (from 8 to 9,7 wks) and increased survival (from 4.7 mth to 6.7 mth, p< 0.001). In that study the biological parameters that correlate with response (mutation status, EGFR gene amplification and EGFR immunohistochemistry) are still under study.

Survival benefit correlated best with the presence of EGFR expression by immunohistochemistry and EGFR gene polysomy (22).

Inclusion Criteria:

• Histological or cytological diagnosis of inoperable, locally advanced, recurrent or metastatic (Stage IIIB or Stage IV) adenocarcinoma of the lung in a patient with a smoking history of < 15 years and quit smoking > 1 year before diagnosis.
• Evidence of disease but measurable disease is not mandatory.
• 18 years of age or older.
• ECOG performance status of 0 - 3.
• Patients not eligible for standard curative-intent treatment with surgery or chemo-radiotherapy.
• Life expectancy ³ 3 months.
• Prior therapy for NSCLC allowed for primary disease: surgery and radiotherapy and adjuvant or proto-adjuvant chemotherapy completed > 6 months before inclusion
• Adequate bone marrow, hepatic and renal function:

Granulocyte count > 1.5 x 109/L and platelet count > 100 x 109/L Serum bilirubin must be < 1.5 upper limit of normal (ULN). If alkaline phosphatase is > 2.5 x ULN, SGOT (AST) and SGPT (ALT) must be < 1.5 x ULN.

Serum creatinine < 1.5 ULN or creatinine clearance > 60 ml/min.

• Ability for giving informed consent for participating in the study and filling out FACT-L quality of life scales.
• Able to comply with study and follow-up procedures.
• Availability of tumour biopsy sample (fixed in formalin and, if possible, also snap frozen tumour sample). If frozen samples are available, these will be collected by central data management.
• Signed Informed Consent for performing mutation analysis and subsequent biomarker analysis.
• Separate signed Informed Consent for participation in the treatment phase of the study.
• Ability to take oral medication.
• For all females of childbearing potential a negative pregnancy test must be obtained within 48 hours before starting therapy.

Exclusion Criteria:

• Patients for whom urgent chemotherapy or radiotherapy is deemed necessary (e.g. rapidly progressive disease).
• Current symptomatic central nervous disorder, brain or leptomeningeal metastasis.
• Pre-existing symptomatic interstitial lung disease, not related to the current malignancy.
• Patients with a history of other malignancies, except patients with basal cell carcinoma of the skin or in situ carcinoma of the cervix with a disease free interval of ³ 5 years. Patients with a prior history of other good prognosis malignancies more than 5 years since end of treatment and in un-maintained complete remission also can be considered for inclusion
• Prior therapy with systemic anti-tumour therapy with HER1/EGFR inhibitors (small molecule or monoclonal antibody) or chemotherapy
• Significant malabsorption syndrome or disease affecting the gastrointestinal tract function
• Pregnant or breast-feeding women; for women in reproductive condition, a negative pregnancy test is required.
• Concomitant food or drug intake which potentially impairs absorption and metabolisation of RTKI's.
• Participation in another clinical trial with any investigational drug within 30 days prior to study screening.
• Any unstable systemic disease (including active infection, grade 4 hypertension, unstable angina, congestive heart failure, hepatic, renal or metabolic disease).
• Any significant ophthalmological abnormality, especially severe dry eye syndrome, keratoconjunctivitis sicca, Sjögren syndrome, severe exposure keratitis or any other disorder likely to increase the risk of corneal epithelial lesions.