Baseline characteristics
From January 2011 to October 2016, 363 patients with advanced and metastatic non-neuroendocrine non-small-cell lung cancer were enrolled based on the inclusion and exclusion criteria. The baseline characteristics are summarized in Table 1. Among these 363 patients, 239 (65.8%) patients were male and 124 (34.2%) were female, and age ranged from 27 to 95 (mean 66.1 ± 11.4). Based on the TNM staging system, 75 patients presented with stage IIIB, followed by 9 with IIIC, 146 with stage IVA, and 133 with stage IVB. Regarding the pathological differentiation level of tumors, 0.6%, 4.9%, and 94.5% of patients were histologically diagnosed with well, moderately, and poorly differentiated disease, respectively. Eighty-seven patients had EGFR mutations, and 276 patients were EGFR-wild type or unknown. The average value of NSE level was 20.1 ng/ml (range 0.1-285.2 ng/ml).
The optimal cut-off value for NSE level
According to the X-tile program, the optimum cut-off value for NSE level was 26.1 ng/ml (Figure 1A-C). Based on the optimal cut-off value, patients were subsequently divided into two groups (NSE < 26.1 ng/ml and NSE ≥ 26.1 ng/ml) for further analyses.
Correlation between NSE level and other clinicopathological characteristics
Potential correlations of NSE levels with other clinicopathological variables were explored (Table 2). The chi-squared test revealed that the NSE level was correlated with the ECOG score (P = 0.003), histological source (P = 0.008), histological type (P = 0.014), and clinical stage (P < 0.001).
Multivariate logistic regression analysis of NSE level and clinical factors showed that NSE level was associated with clinical stage (P = 0.006). No significant difference in the other clinicopathological variables with NSE level was found (Table 3).
Effectiveness of first-line treatment
According to RECIST 1.1 guidlines, among the 363 patients, the outcomes were CR: 0.0% (0/363); PR: 16.0% (58/363); SD: 25.6% (93/363); PD: 58.4% (212/363); ORR: 16.0% (58/363); and DCR: 41.6% (151/363) (Table 4A).
In addition, in the low NSE group (NSE < 26.1 ng/ml), the outcomes were CR: 0.0% (0/298); PR: 18.8% (56/298); SD: 27.2% (81/298); PD: 54.0% (161/298); ORR: 18.8% (56/298); and DCR: 46.0% (137/298). In the high NSE group (NSE ≥ 26.1 ng/ml), the outcomes were CR: 0.0% (0/65); PR: 3.1% (2/65); SD: 18.5% (12/65); PD: 78.5% (51/65); ORR: 3.1% (2/65); and DCR: 21.6% (14/65) (Table 4B).
The outcome of PFS and OS
Among 363 patients, at the end of follow-up, 199 (54.8%) patients had been confirmed dead.
The median PFS was 7.96 months (95% confidence interval [CI] = 6.42-9.50). The 1- and 3-year PFS rates were 40.4% and 9.0%, respectively (Figure 2A). The median OS was 14.41 months (95% CI = 11.83-16.98). The 1- year and 3-year OS rates were 56.4% and 16.6%, respectively (Figure 2B).
The median PFS was 9.93 months (95% CI = 7.41-12.46) in the low NSE group (< 26.1 ng/ml). The mPFS was 4.31 months (95% CI = 2.20-6.42) in the high NSE group (NSE ≥ 26.1 ng/ml) (P < 0.001, log-rank test).
The 1-, and 3-year PFS rates of the low NSE group (NSE < 26.1 ng/ml) were 44.9% and 11.7%, respectively. The 1-year and 3-year PFS rates of the high NSE group (NSE ≥ 26.1 ng/ml) were 14.5% and 0.0%, respectively (Figure 3A). A significant difference in PFS was observed between the two groups (P < 0.001, log-rank test).
The median OS was 17.11 months (95% CI = 13.71-20.50) in the low NSE group (<26.1 ng/ml). The mOS was 8.13 months (95% CI = 6.23-9.97) in the high NSE group (NSE ≥ 26.1 ng/ml)(P < 0.001, log-rank test).
The 1-year and 3-year OS rates of the low NSE group (NSE < 26.1 ng/ml) were 63.6% and 18.6%, respectively. The 1-year and 3-year OS rates of the high NSE group (NSE ≥ 26.1 ng/ml) were 22.6% and 0.0%, respectively (Figure 3B). A significant difference in OS was observed between the two groups (P < 0.001, log-rank test).
The total number of EGFR mutations in adenocarcinoma patients was 87. In this subgroup, patients with NSE < 26.1 ng/ml, mPFS was 16.02 months (95% CI = 10.47-21.57), and the 1-year and 3-year PFS rates were 66.7% and 18.7%, respectively. For patients with NSE ≥ 26.1 ng/ml, the mPFS was 6.58 months (95% CI = 2.36-10.80), and the 1-year and 3-year PFS rates were 13.2% and 0%, respectively. A significant difference in PFS was observed between the two groups (P = 0.0016, log-rank test) (Figure 4A).
The median overall survival of patients with EGFR mutations and low serum NSE levels (< 26.1 ng/ml) was 27.73 months (95% CI = 11.52-15.59). The 1- and 3-year OS rates were 77.7% and 21.3%, respectively. The median overall survival of patients with EGFR wild-type and high serum NSE levels (≥ 26.1 ng/ml) was 9.51 months (95% CI= 3.53-15.48). The 1- and 3-year OS rates were 21.9% and 0%, respectively. A significant difference in OS was observed between the two groups (P < 0.001, log-rank test) (Figure 4B).
There were 171 cases of EGFR-wild type adenocarcinomas. In this subgroup, patients with NSE < 26.1 ng/ml, mPFS was 7.46 months (95% CI = 6.14-8.80), and 1-year and 3-year PFS rates were 29.8% and 0%, respectively. For patients with NSE ≥ 26.1 ng/ml, the mPFS was 3.42 months (95% CI = 2.11-4.73), and the 1-year, 3-year PFS rates were 11.1% and 0%, respectively. A significant difference in PFS was observed between the two groups (P < 0.001, log-rank test) (Figure 4C).
The median overall survival of patients with EGFR-wild type and low serum NSE levels (< 26.1 ng/ml) was 13.55 months (95% CI= 11.52-15.59). The 1- and 3-year OS rates were 56.0% and 13.8%, respectively. The median overall survival of patients with EGFR mutations and high serum NSE levels (≥ 26.1 ng/ml)) was 7.76 months (95% CI= 5.79-9.74). The 1- and 3-year OS rates were 18.9% and 5.1%, respectively. A significant difference in OS was observed between the two groups (P < 0.001, log-rank test) (Figure 4D).
In the patients with serum NSE < 26.1 ng/ml, the cumulative risk of brain metastasis at 3 months, 6 months and 9 months was 11.4%, 26.7% and 50.3%, respectively.
In the patients with serum NSE ≥ 26.1 ng/ml, the cumulative risk of brain metastasis at 3 months, 6 months and 9 months was 55.4%, 79.2% and 97%, respectively. There was a significant difference between the two groups (P < 0.001, log-rank test) (Figure 5).
Univariate and multivariate analysis of PFS
As shown in Figure 6A, the results from univariate analysis indicated that gender (HR = 0.72, P = 0.028), smoking history (HR = 1.44, P = 0.013), clinical stage (HR = 2.58, P < 0.001), pathological differentiation (HR = 5.23, P < 0.001), NSE level (HR = 2.40, P < 0.001), and EGFR mutation status (HR = 0.58,P = 0.001) were significant prognostic factors for PFS. Multivariate analysis indicated that clinical stage (HR = 1.93, P = 0.001), pathological differentiation (HR = 3.24, P = 0.007 ), NSE level (HR = 1.81, P = 0.001), and EGFR mutation status (HR = 0.54, P < 0.001) were independent prognostic parameters for PFS (Figure 6B).
Univariate and multivariate analysis of OS
Univariate analysis in all patients showed that gender (HR = 0.67, P = 0.008), smoking history (HR = 1.46, P = 0.010), clinical stage (HR = 2.34, P < 0.001), pathological differentiation (HR = 4.59, P < 0.001), NSE level (HR = 2.40, P < 0.001), and EGFR mutation status (HR = 0.57, P = 0.001), were associated with OS (Figure 7A). To identify possible independent prognostic factors, multivariate OS analysis next was conducted in which ECOG score (HR = 1.33, P = 0.016), clinical stage (HR = 1.94, P = 0.001), pathological differentiation (HR = 2.85, P = 0.015), NSE level (HR = 1.76, P = 0.002), and EGFR mutation status (HR = 1.58, P = 0.002) were identified as independent factors predicting OS (Figure 7B).