Efficacy of First-line Chemotherapy Affects the Second-Line Setting Response in Patients with Advanced Non-Small Cell Lung Cancer

Lung cancer remains the leading cause of cancerrelated deaths worldwide (Ramalingam et al., 2011). Nonsmall cell lung cancer (NSCLC) accounts for about 80% of lung cancer cases and most are diagnosed at advanced stages (Lal et al., 2011; Kirmani et al., 2013). Tailored therapy based on biomarker analysis has entered the reality of lung cancer treatment (Qi et al., 2012; Gridelli et al., 2014). Patients with EGFR activated mutation or ALK/ ROS1 fusion benefit from targeted therapy with EGFR TKIs or ALK inhibitors (Gao et al., 2012). However, only a minority of patients express these markers, with EGFR mutations detected in about 30-40% and EML4-ALK fusions in about 4% in an East Asian population (Zhou et al., 2011; Ren et al., 2012). So for the majority of patients without driver mutations, platinum-based chemotherapy remains the standard first-line treatment (Natukula et al.,


Introduction
Lung cancer remains the leading cause of cancerrelated deaths worldwide (Ramalingam et al., 2011). Nonsmall cell lung cancer (NSCLC) accounts for about 80% of lung cancer cases and most are diagnosed at advanced stages (Lal et al., 2011;Kirmani et al., 2013). Tailored therapy based on biomarker analysis has entered the reality of lung cancer treatment (Qi et al., 2012;Gridelli et al., 2014). Patients with EGFR activated mutation or ALK/ ROS1 fusion benefit from targeted therapy with EGFR TKIs or ALK inhibitors (Gao et al., 2012). However, only a minority of patients express these markers, with EGFR mutations detected in about 30-40% and EML4-ALK fusions in about 4% in an East Asian population (Zhou et al., 2011;. So for the majority of patients without driver mutations, platinum-based chemotherapy remains the standard first-line treatment (Natukula et al.,
In this subpopulation, patients are recommended to receive second-line chemotherapy such as doctacel or pemetrexed when first-line treatment fails (Hanna et al., 2004;Caponi et al., 2010;Ardizzoni et al., 2012) However, the efficacy of second-line chemotherapy is still dismal, with a response rate of about 10%, PFS of 2-3 months and OS of 6-9 months (Hanna et al., 2004;de Marinis et al., 2011;Ardizzoni et al., 2012). More importantly, currently, there are also still no identified predictive factors to predict the efficacy of second-line chemotherapy (Favaretto et al., 2009;de Marinis et al., 2011).
To better clarify the potential predictive factors for the outcome of second-line chemotherapy, we performed a retrospective study to investigate the association of clinical pathological features such as the efficacy of first-line chemotherapy, histology, gender and ECOG PS in 1008 Chinese patients with advanced NSCLC .

Materials and Methods
A total of 3520 patients with histologically confirmed primary stage IIIB or IV NSCLC received platinum-based doublets for first-line therapy in Shanghai Pulmonary Hospital, China from September 2005 to July 2010. Of these, 1008 patients (28.6%) with evidence of progressive disease within 6 months after the end of first-line treatment who received second-line chemotherapy were included in our study. These 1008 patients were classified according to response to first-line chemotherapy and period of progression ( Figure 1). These patients were followed up until January 1, 2013, and 784 deaths were recorded. Parameters such as efficacy of first-line chemotherapy, period of progression, gender, age, smoking history, ECOG PS, histological type, and metastasis location at diagnosis (liver, bone and brain) were chosen as potential prognostic factors for analysis in our study.

Tumor evaluation and response
Tumor response was evaluated every 6-8 weeks according to the Response Evaluation Criteria in Solid Tumors (RECIST1.1) (Nishino et al., 2010). PFS was calculated as the time between the first day of treatment and the time to documented disease progression, the last follow-up data or death from any cause. OS was calculated as the time from the start of second-line therapy to death or the last follow-up data (Zietemann et al., 2011;Di Maio et al., 2012).

Statistical analysis
The association between the efficacy of first-line treatment or related baseline factors and ORR for secondline treatment was assessed using the Chi-square test (Ardizzoni et al., 2012). The Kaplan-Meier method was used to estimate PFS and OS and the log-rank test was used to compare survival curves (Di Maio et al., 2012). A multivariate Cox-regression model was used to analyze prognostic factors for PFS and OS, presenting a hazard ratio (HR) and confidence interval (CI) for the independent variables. Statistical analyses were performed using SPSS 17.0, and a p-value less than 0.05 was considered statistically significant (Vergnenegre et al., 2011).

The efficacy of first-line chemotherapy and its effect on second-line chemotherapy
The ORR and disease control rate were 25.8% and 71.4% for the first line setting, while they were 9.4% and 41.4% for the second-line chemotherapy in the 1008 patients, respectively.
The ORR for second-line therapy was 15.4% in patients who responded for first-line therapy, which was significantly higher than 10.1% in these with SD and 2.3% in these with PD (p<0.001). The ORR were also significantly different according to the period of progression, which were 2.4%, 8.8% and 18.7% in patients who progressed on first-line therapy, within 3 months and between 3 to 6 months after the last dose of first line chemotherapy respectively ( p<0.001).

Survival analysis for second-line chemotherapy
The median PFS was 3.80 months and 2.44 months for first-line and second-line chemotherapy respectively. The median PFS and OS of second-line therapy differed significantly according to the different response status to first-line therapy. The median PFS were 3.80, 2.77 and 2.03 months in patients who response, got a stable disease and progressed from the first line chemotherapy (p<0.001), while the median OS were 11.60 10.33 and 6.57 months respectively according to the different response status (p<0.001) (Figure 2). In terms of period  of progression, significant differences were also found on PFS and OS. The median PFS were 2.13, 2.63 and 4.20 months in patients who progressed on first-line therapy, within 3 months and between 3 to 6 months after the last dose of first line chemotherapy respectively (p<0.001), while the median OS were 6.97, 9.57 and 13.1 months respectively according to different period of progression (p<0.001) (Figure 3). Never smokers, patients with ECOG PS 0-1, female gender, non-squamous cell carcinoma and those without liver metastasis at diagnosis had longer PFS ( . 1.93 months, p<0.001). However, no differences were found between patients with and without brain or bone metastasis, or between elderly and young patients ( Table 2)As for OS, patients of never smokers and female had a longer median OS after second-line therapy (11.00 vs. 7.77 months, p<0.001; 12.93 vs. 8.77 months, p<0.001). Also, patients with nonsquamous cell carcinoma or ECOG PS 0-1 had a longer median OS than those with squamous cell carcinoma or PS 2 (10.33 vs. 6.97 months, p=0.001; 9.80 vs. 6.43 months, p=0.006). Patients with liver metastasis at baseline had a worse median OS after second-line therapy (6.13 vs. 9.80 months, p<0.001). However, no differences were found between elderly and young patients, or patients with and without brain or bone metastasis (Table 2).

Cox-regression model of the possible prognostic factors for PFS and OS after second-line chemotherapy
The following characteristics significantly influenced PFS after second-line therapy both by univariate and multivariate analysis: histology, period of progression, liver metastasis, the efficacy of first-line chemotherapy and gender (Table 3). Smoking status and ECOG PS were significant only by univariate analysis.

Discussion
In this study, we retrospectively analyses the prognosis of second-line chemotherapy in a total of 1008 patients. We found that the variables such as gender, age, smoking status, ECOG PS, liver metastasis, best efficacy of firstline therapy, and time of progression were associated with survival outcome. Among them, without liver metastasis, never smoking status, ECOG PS 0-1, non-squamous cell carcinoma, response to first-line therapy, and progression within 3-6 months after first-line chemotherapy were statistically significant associated with longer OS in multivariate analyses after second-line chemotherapy.
In terms of clinical characteristics, several previous studies have shown that different histology, smoking status, and ECOG PS could have a significantly different overall survival (Tadashi et al., 2000;Sun et al., 2010;Di Maio et al., 2012;Ali et al., 2012). In 2010, Jong-Mu Sun et al. found that female gender, adenocarcinoma, never smoking status, and ECOG PS of 0-1 were good predictive factors for subsequent pemetrexed therapy after prior gemcitabine-based chemotherapy in advanced NSCLC (Sun et al., 2010). Consistent with these reports, our study found ECOG PS 0-1 significantly associated with longer PFS and OS and numerically higher ORR. Also, patients of non-squamous cell carcinoma, never smoking status and female showed a significantly longer PFS and OS after receiving second-line chemotherapy in the univariate analysis, while female patients did not remain a predictor for longer PFS and OS in the multi-variate analysis. The EGFR mutation distribution in different subgroup might attribute to this result (Wu et al., 2010). As we all knows, it were histological type and smoking status but not gender significantly affect the incidence of EGFR mutation. Furthermore, patients with EGFR mutation showed a superiority to the efficacy of chemotherapy (Pao et al., 2010;Gridelli et al., 2014).
As far as we know, this study is the first one to comprehensively analysis the efficacy of first-line chemotherapy affects the response of second-line setting in 1008 patients with advanced non-small cell lung cancer. We found that best response to first-line therapy a good predictor for not only OS but also ORR and PFS following second-line chemotherapy, and the trend also existed in multivariate analysis. Furthermore, our study is the first to identify that a longer progression period as an independent positive prognostic factor for OS after second-line treatment. In line with our study, the results from several studies with limited samples suggested that the efficacy of first-line chemotherapy could affect secondline outcomes (Weiss et al., 2007;Moro-Sibilot et al., 2010;Greillier et al., 2012;Ali et al., 2012). Also, in the JMEI trial which compared pemetrexed with docetaxel as second-line setting in patients advanced NSCLC, Hanna et al found a trend that patients who had a clinical benefit from first-line therapy were more likely to benefit from second-line therapy (Hanna et al., 2004). The findings from our study, together with these previous studies suggested that the efficacy of the first line chemotherapy could be helpful to choose suitable patients for secondline chemotherapy.
Besides that, liver metastasis was known as a poor prognosis for several kinds of cancers (Zabaleta et al., 2011;Gomez et al., 2012;Ishizuka et al., 2012). Tadashi Maeda et al. found that liver metastasis was an independent prognostic factor for first-line chemotherapy in advanced NSCLC (Tadashi et al., 2000). To our knowledge, this study is the first one to investigate the effect of liver metastasis on efficacy of second-line setting in patients with NSCLC We found that liver metastasis was a strong poor predictor on PFS and OS after secondline chemotherapy. More work is needed to improve the prognosis in the NSCLC patients with liver metastasis.
Asian Pacific Journal of Cancer Prevention, Vol 15, 2014 6803 DOI:http://dx.doi.org/10.7314/APJCP.2014.15.16.6799 Efficacy of First-line Chemotherapy Affects Second-Line Setting Response in NSCLC Patients There are several limitations in this study. Firstly, we did not further analyze which might be the most beneficial agents (pemetrexed, docetaxel, or others) for patients with disease progression. Secondly, the interaction between first-and second-line drugs should also be included. Besides, for analysis of prognostic factors on efficacy of second-line treatment, we did not consider molecular markers such as TS expression, BRCA1 expression, EGFR mutation or ALK fusion, which have established roles both as determinants of treatment choice and prognostic factors.
In conclusion, patients who responded to first-line treatment had a better outcome after second-line therapy for advanced NSCLC than those who did not respond. Furthermore, the period of progression, histology, liver metastasis, smoking status and ECOG PS were independent prognostic factors for PFS and OS following second-line chemotherapy. These findings could help better evaluate the risks and benefits associated with therapeutic decisions and be used to select patients optimally for the proper choice of treatment in physicians' daily clinical practice.