ESMO 2014: How to Guide the Selection of Chemotherapy Regimen for Non-Squamous Non-Small Cell Lung Cancer

Results from two Korean studies of docetaxel plus cisplatin vs pemetrexed plus cisplatin or pemetrexed plus cisplatin vs gemcitabine plus cisplatin

Although docetaxel plus cisplatin vs pemetrexed plus cisplatin showed similar progression-free survival (PFS) and response rate (RR) in a phase III study of first-line treatment for non-squamous non-small cell lung cancer (NSCLC), more frequent adverse events and higher toxicities were observed in the docetaxel arm. In another Korean, biomarker-stratified, randomised phase II study, RR and PFS with pemetrexed/cisplatin regimen compared to gemcitabine/cisplatin chemotherapy were more prominent in the thymidylate synthase (TS)-negative group than in TS-positive group, suggesting that TS can be used as a predictive biomarker. The results from both studies were presented during the Proffered Paper session in metastatic NSCLC at ESMO Congress 2014 in Madrid, Spain.

Docetaxel plus cisplatin vs pemetrexed plus cisplatinin first line non-squamous NSCLC

For patients with non-squamous-NSCLC, pemetrexed plus cisplatin is superior to gemcitabine plus cisplatin in terms of efficacy and toxicity. Docetaxel/cisplatin is an active regimen for first- line NSCLC. Median PFS of pemetrexed/cisplatin is 6.4 months in East Asian compared with 5.3 months in all ethnic patients. However, there was no prospective phase III trial that directly compared the efficacy of two regimens. The objective of the study presented by Prof. Young-Chul Kim of the Pulmonology Unit, Chonnam National University Hwasun Hospital Lung Cancer Clinic, Hwasun Gun, Korea was to prove the non-inferiority of PFS of docetaxel/cisplatin compared with pemetrexed/cisplatin.

The researchers performed an open-label phase III trial. The study recruitment was between August 2011 and December 2013 at 14 centers in Korea. Patients with chemotherapy-naïve non-squamous-NSCLC were randomised into 3 weekly cisplatin-based chemotherapy, with either docetaxel or pemetrexed, for up to 4 cycles with stratification by performance status and sex. Thereafter, the patients continued treatment with either emetrexed, EGFR tyrosine kinase inhibitor or docetaxel.

The primary objecive was to assess PFS and the secondary endpoints were RR assessed by RECIST 1.1 criteria, overall survival (OS) and safety.

In total, 156 patients were randomised, but after 149 patients included into pemetrexed/cisplatin  (77) and docetaxel/cisplatin (72) arms, the study team finished enrolment because of approval and popular use of maintenance treatment with pemetrexed in Korea.

Clinical characteristics including sex, age, and performance status were well balanced between the arms. Number of cycles and relative dose intensity were not significantly different between the arms.

In intent to treat population, partial response was observed in 24 (31.2%) and 24 (33.3%) patients in pemetrexed/cisplatin and docetaxe/cisplatin group, respectively.

Median PFS was 4.7 months in the pemetrexed/cisplatin arm and 4.6 months in the docetaxel/cisplatin arm. Median OS was 19.7 month in the pemetrexed/cisplatin arm and 28.0 month in the docetaxel/cisplatin arm.

Rate of grade 3 or 4 neutropaenia and febrile neutropaenia, number of serious adverse events were significantly higher in docetaxel arm.


Prof. Kim concluded that in non-squamous NSCLC without driver mutations, both regimens showed similar PFS and RR. However, more frequent adverse events and higher toxicities were observed in the docetaxel/cisplatin arm. Numerically shorter PFS in both arms in this trial in comparison to the East Asian and overall population in other trials suggest that maintenance treatment should be considered unless disease progression is noted.

Dr Giorgia Scagliotti who discussed the study results said that it was planned as a non-inferiority study with a margin of 1.5 months. The original sample size was 562 patients, however, it was stopped after enrollment of 159 patients (28%) and for him this is just a randomised phase 2 study, therefore no meaningful conclusion could be drawn because of the limited number of patients. The PFS data are comparable to Caucasian patients, longer OS data in both arms are typical for data seen in Asian patients. Lung toxicity is proper higher for docetaxel/cisplatin treatment. Pemetrexed-based chemotherapy remains the preferred doublet in non-oncogene addicted non-squamous NSCLC.

Pemetrexed plus cisplatin vs gemcitabine plus cisplatin according to thymidylate synthase expression in non-squamous NSCLC

Prof. Myung-Ju Ahn of the Department Of Internal Medicine, Hematology, Oncology, Sungkyunkwan University, School of Medicine, Seoul, Korea presented results of a  biomarker-stratified randomised phase II trial. The primary endpoint was to determine the predictive value of TS in non-squamous NSCLC in term of objective response rate by testing the interaction between treatment arm and TS positivity.

As background, the Korean researchers explained that pemetrexed/cisplatin showed superior outcomes in non-squamous NSCLC compared with gemcitabine/cisplatin while it was inferior in squamous NSCLC in a phase III trial. One of culprits for this phenomenon is the different expression level of TS according to histotypes where TS expression is higher in squamous NSCLC than in non-squamous NSCLC.

TS is an important enzyme inde novoDNA synthesis and one of target proteins of pemetrexed. Many retrospective analyses showed better clinical outcomes of pemetrexed-based therapy in NSCLC patients with low TS-expressing tumours. This study was conducted to evaluate whether TS expression is a predictive factor for pemetrexed/cisplatin compared with gemcitabine/cisplatin in non-squamous NSCLC patients.

The main inclusion criteria were patients ≥ 18 years with histologically or cytologically confirmed advanced stage (IIIB or IV, or recurrent disease at least 6 months after complete resection), non-squamous NSCLC except large cell neuroendocrine carcinoma, at least one measurable lesions, ECOG performance status 0 or 1, adequate organ function, and no prior chemotherapy for advanced disease. Written consent was obtained from eligible patients.

TS expression was measured by immunohistochemistry. The patients with more than 10% of tumour expressing TS were grouped as a TS-positive and those with 10% or less were grouped as a TS-negative. In a retrospective study of the same investigators, the median H score of 15 as the cut-off value for TS-positive or –negative tumours was identical to the expression in 10% of tumour cells irrespective of staining intensity.

After being stratified as TS-positive or TS-negative, patients were randomised to either pemetrexed/cisplatin or gemcitabine/cisplatin arms by 1:1 fashion.

The study primary endpoint was interaction between TS and treatment allocation in term of response rate. Secondary endpoints included interaction between TS and treatment allocation in terms of PFS or OS, and safety of treatment.

This trial was designed to provide 90% power to detect an interaction term of b=1.294 based on the response rate of each group as estimated from the investigators’ previous study. One-sided alpha of 0.1 was used for the analysis of interaction term, whereas two-sided alpha of 0.05 was used for other endpoints.  

Chemotherapy in both arms was administered until disease progression or unacceptable toxicity with maximum 6 cycles. Response evaluation by RECIST 1.1 was done every 2 cycles during treatment and every 2 months after completion of study chemotherapy.

For 315 evaluable patients, the RR calculated by investigators in pemetrexed/cisplatin and gemcitabine/cisplatin were 47.0% and 21.1% in TS-negative group, whereas 40.3% and 39.2% in TS-positive group (p = 0.008). The RR in pemetrexed/cisplatin and gemcitabine/cisplatin arms evaluated by independent reviewers were 38.6% and 21.1% in TS-negative group, and 40.3% and 48.1% in TS-positive group (p = 0.007).

The median PFS in pemetrexed/cisplatin and gemcitabine/cisplatin arms were 6.4 and 5.5 months in TS-negative group (p = 0.013), whereas 5.9 and 5.3 months in TS-positive group (p = 0.64) (interaction p = 0.07).

Out of 88 patients with tumours harboring EGFR mutations, 74 received gefitinib or erlotinib, while 9 were only observed without any post-discontinuation therapy because their diseases did not yet progress at data cut-off time.

The median OS in pemetrexed/cisplatin and gemcitabine/cisplatin arms were not different in TS-negative group (not reached vs. 28.3 months, p = 0.86), as well as in TS-positive group (19.0 vs. 14.4 months, p = 0.36) (interaction p = 0.31). However, in multivariate analyses, TS-negative expression was significantly associated with longer survival (HR 0.64), along with younger age (HR 0.62) and EGFR mutation (HR 0.45).


The authors concluded that in term of RR and PFS, clinical benefits with pemetrexed/cisplatin chemotherapy compared with gemcitabine/cisplatin were more prominent in the TS-negative group than in the TS-positive group, suggesting that TS can be used as a predictive biomarker. Furthermore, given that low TS expression was associated with prolonged OS irrespective of chemotherapeutic regimens, the authors noted that TS expression can also serve as a prognostic biomarker.

Dr Giorgio Scagliotti who discussed the study results said that it was a reasonably well designed study, however, prognostic/predictive role of TS remains unsolved. Immunohistochemistry, according to Dr Scagliotti, should not be used, while a role of RT-PCR detection would be more appropriate. Pharmacogenomic markers are still reserved to clinical research settings and no data support the use in clinical practice to select patiets.


The LBA41_PR study was funded by Sanofi Aventis Korea. All study authors of LBA42_PR have declared no conflicts of interest.