Introduction

Lung cancer is amongst the major causes of cancer-related mortality, globally [1]. Non-small cell lung cancer (NSCLC) accounts for approximately 85% of all the lung cancer cases [2]. Targeted therapies are actively being developed for treatment of select patients with NSCLC. Tyrosine kinase inhibitors (TKIs), such as gefitinib and erlotinib, that target the epidermal growth factor receptor (EGFR), are small molecule drugs introduced in the clinics for the treatment of NSCLC patients. NSCLC patients with EGFR mutations who received TKI such as gefitinib, erlotinib, afatinib, and osimertinib had better progression free survival (PFS) and response rates than patients who received chemotherapy using cytotoxic drugs [3,4,5,6,7]. Based on these results, EGFR-TKI has become a standard treatment regimen for patients with advanced NSCLC harboring EGFR mutations. However, 20–30% of NSCLC cases show primary resistance to EGFR-TKIs, despite harboring an activating EGFR mutation [8]. Previously, we have reported the patient’s smoking status as an independent predictor of response and PFS upon treatment with EGFR-TKI therapy [9, 10]. However, other indicators that can elaborate on the response to EGFR-TKI therapy remain largely understudied.

Complete blood count (CBC) is one of the most common laboratory tests performed in the clinics. The absolute counts of neutrophils and lymphocytes reflect the inflammatory response and overall immune status of the patients. Previous studies showed that the peripheral blood prognostic inflammatory markers such as the neutrophil-to-lymphocyte ratio (NLR) associated with patient’s prognosis and treatment outcome [11,12,13,14]. However, there are a limited number of reports about the relationship between these inflammatory markers and the efficacy of EGFR-TKIs in advanced NSCLC patients with EGFR mutations.

Here, we conducted a clinical study to evaluate the potential of NLR obtained from CBCs of patients with advanced NSCLC harboring EGFR mutations in predicting the clinical efficacy of treatment with EGFR-TKIs.

Materials and methods

Patient selection and data collection

For the retrospective analysis, a total of 205 patients with advanced NSCLC who received EGFR-TKIs, including gefitinib and erlotinib, at the Kitasato University Hospital (Kanagawa, Japan) between March 2009 and June 2016, were enrolled. The date cut-off date was March 2019. Patients with histologically or cytologically confirmed NSCLC, stage IV disease, or postoperative recurrence (according to the criteria of the Union for International Cancer Control, version 7), and those not suitable for curative treatment, were assessed for patient selection. Consecutively, patients that met the following inclusion criteria were eligible for the study: (1) measurable target lesions observed in the chest X-ray, computed tomography of the chest and abdomen, or by other imaging modalities [magnetic resonance imaging (MRI) of the head, positron emission tomography (PET), or positron emission tomography/computed tomography (PET/CT)]; and (2) histologically confirmed NSCLC. Furthermore, the patients were categorized according to the smoking history as current smokers (including former smokers who are not categorized former light smokers), former light smokers (defined as patients who had stopped smoking at least 15 years previously, with a total of ≤10 pack-years of smoking), and never smokers (defined as patients who had smoked <100 cigarettes in their lifetime). The CBC was tested before the EGFR-TKI treatment, and NLR was calculated based on the absolute neutrophil and lymphocyte counts. The ethical review board committee of the Kitasato University and its affiliated hospitals approved the present study, which received ethical approval for the use of an opt-out style.

Response assessment

Tumor response was classified on the basis of the Response Evaluation Criteria for Solid Tumors, version 1.1. The disease stage prior to EGFR-TKI therapy and disease progression or recurrence were determined by physical examination, chest X-ray, CT of the chest and abdomen, or by other imaging modalities (e.g., MRI of the head and PET scan).

Statistical analyses

NLR was used to compare patient characteristics with response rates using the Chi-square test. PFS was measured from the date of commencing EGFR-TKI therapy till the date of disease progression, death, or last follow-up. OS was calculated from the date of commencing EGFR-TKI therapy till death from any cause. Survival curves were plotted using the Kaplan-Meier method. Log-rank test was used to assess differences between PFS and OS based on NLR for each patient. Variables (including gender, age, performance status [PS], NLR, EGFR genotype, smoking status, clinical stage prior to EGFR-TKI treatment [Stage IV vs. postoperative recurrence], the presence or absence of brain metastases, type of EGFR-TKI [gefitinib vs. erlotinib]) were entered into a Cox proportional hazards regression model to estimate the hazards ratios for PFS and OS. The receiver operating characteristic (ROC) curves and Youden’s index were utilized to determine the optimal cut-off for NLR. Results with P-values <0.05 were considered statistically significant. All statistical analyses were performed using SPSS software for Windows, version 23.0 (IBM Corp., Armonk, NY, USA).

The statistical significance of differences in the NLR according to the response to EGFR-TKI was determined by the Welch’s t test.

Results

Correlation of NLR with NSCLC patient characteristics

The clinical characteristics of the patients have been listed in Table 1. The patient cohort comprised of 60% women, with 70 years median age, and 65% of good PS. Ninety seven percent of the patients presented with lung adenocarcinoma (198 patients), while others were either pathology unspecified (3%). All patients were confirmed with stage IV disease or postoperative recurrence. The number of Stage IV patients and postoperative recurrences were 154 and 51 of the 205, respectively.

Table 1 Patient characteristics
Full size table

Cut-off values for immunologic parameters

We used PFS longer or shorter than 10 months as the binary variables for ROC curves [15]. Based on the highest Youden index (specificity+sensitivity–1), an optimal cut-off value of 3.55 was chosen for NLR, with an area under the curve (AUC) value of 0.67 [95% confidence interval (CI): 0.59–0.74, P < 0.0001] (Fig. 1). Furthermore, the comparative analysis for categorical variables suggested high NLR levels in stage IV than in recurrence group (P = 0.02, Table 1).

Fig. 1
figure 1

Receiver operating characteristic curves analysis for NLR in all patients

Full size image

NLR predicts response to EGFR-TKI

An objective response was achieved in 124 of the 205 patients, indicating a 60.5% overall response rate (95% CI: 53.8–67.2%, Table 2). Furthermore, a statistically significant difference (P = 0.009) was observed upon comparison between the low-NLR group (NLR < 3.55) with 69.2% response rate (95% CI: 60.3–78.1%) than the high-NLR group (NLR ≥3.55) with 51.5% rate (95% CI: 41.8–61.2%). In addition, the mean values of NLR in responders (patients with partial response) and non-responders (patients with stable disease or progressive disease) were 4.8 and 11.3 respectively, indicating a significant difference in the mean NLR between them (P = 0.021, Fig. 2).

Table 2 Response rates according to NLR
Full size table
Fig. 2
figure 2

An evaluation of NLR in NSCLC patients with EGFR mutation before the EGFR-TKI therapy according to the response to the EGFR-TKI

Full size image

Survival analysis of NSCLC patients

The patients presented with a median 25.2 months follow-up period for the survival analysis. The median PFS and OS of all the patients together were 10.8 months (95% CI: 8.8–12.8 months, Fig. 3a) and 28.0 months (95% CI: 23.5–32.5 months, Fig. 3b), respectively. The median PFS for the low-NLR and high-NLR group was 15.7 months (95% CI: 12.7–18.7 months) and 6.7 months (95% CI: 4.6–8.8 months, Fig. 4a), respectively (P < 0.0001). The median OS for the low-NLR and high-NLR group was 37.6 months (95% CI: 30.3–44.9 months) and 19.2 months (95% CI: 10.6–27.8 months), respectively (P < 0.001, Fig. 4b). The multivariate analyses commonly identified PS, NLR, smoking status, and stage as significant and independent predictors of PFS, as summarized in Table 3. Moreover, the PS and NLR were identified as independent predictors of OS based on the multivariate analyses (Table 4). Additionally, the never or former light smokers/low-NLR group and the smokers/high-NLR group had median PFS of 16.2 months (95% CI: 13.2–19.2 months) and 5.3 months (95% CI: 1.4–9.2 months), respectively (P = 0.0002, Fig. 5).

Fig. 3
figure 3

Kaplan-Meier plots of a PFS and b OS for all patients

Full size image
Fig. 4
figure 4

Kaplan-Meier plots of a PFS and b OS according to NLR

Full size image
Table 3 Multivariate analysis of prognostic factors for progression-free survival
Full size table
Table 4 Multivariate analysis of prognostic factors for overall survival
Full size table
Fig. 5
figure 5

Comparison of PFS between the never or former light smokers with low-NLR group and smokers with high-NLR group

Full size image

Discussion

EGFR-TKIs are primary treatment for the advanced stage NSCLC patients with EGFR mutations [3,4,5,6,7,8]. However, a significant percentage of NSCLC patients positive for EGFR mutations show resistance to EGFR-TKIs with short-term PFS. Previously, our study suggested that never-smokers and postoperative recurrence statuses serve as predictors for response to EGFR-TKIs, such as gefitinib and erlotinib, and PFS and OS in patients with NSCLC harboring activating EGFR mutations [9, 10]. Here, low NLRs at baseline were significantly associated with favorable tumor response and better PFS and OS in EGFR-mutant NSCLC patients treated with EGFR-TKIs than in patients with high NLRs. Moreover, our analysis, with a significantly larger patient cohort, is consistent with other studies investigating the correlation between NLR and efficacy of treatment with EGFR-TKI for NSCLC patients with EGFR mutations (Table 5) [15,16,17,18,19,20,21,22]. Furthermore, our NLR cut-off value of 3.55 is in the range reported by others (NLR 2.1–5.0) to predict the clinical efficacy for the treatment of NSCLC patients harboring TKI-sensitive EGFR mutations [15,16,17,18,19,20,21,22].

Table 5 Previous reports investigating the correlation between NLR and efficacy of EGFR-TKI for NSCLC patients with EGFR mutation
Full size table

The utility of NLR as a predictive factor in cancer patients remains relatively understudied. Growing evidences indicate molecular and cellular pathways involving inflammation that contribute to proliferation, angiogenesis, and metastasis of neoplastic cells [23, 24]. Moreover, in patients with mesothelioma, NLR ≥5 correlated with elevated expression of Ki-67 and vascular endothelial growth factor, indicating increased tumor cell proliferation and sustained angiogenesis, than in patients with NLR <5 [25]. Furthermore, the circulating neutrophils release diverse inflammatory cytokines, including tumor necrosis factor-α and interleukin-6, leading to cancer progression [26].

The tumor microenvironment (TME) is involved in tumorigenesis and malignant progression and has a major impact on both the therapeutic response of the tumor and therapeutic effectiveness [27, 28]. Tumor-infiltrating lymphocytes (TILs) are important components of the TME that regulate inflammatory responses and play a pivotal role in eradicating tumor cells [29]. In TME, a significant increase in TIL (CD-8+ T-cells) and decrease in regulatory T-cells have been observed upon administration of EGFR-TKI in a lung cancer mouse model harboring EGFR mutation [30]. Thus, it may be reasonable to argue that treatment with EGFR-TKI is more effective in NSCLC patients positive for EGFR mutations with low NLR than in those with high NLR.

Next, our analysis suggests that NLR can be correlated with the PFS of NSCLC patients, where never or former light smokers showed better survival than the smokers. Furthermore, a combination of pretreatment NLR and smoking status served well as predictors of response to EGFR-TKI. Taken together, the analysis suggests that EGFR-mutant NSCLC patients with high NLR and smoking history show poor survival outcomes when treated only with EGFR-TKI, indicating the necessity of an alternative or combined chemotherapy to provide better response and survival outcome in the patients.

However, our study has certain limitations. First, being a retrospective study, the result cannot be regarded as definitive. Second, we could not exclude the involvement of other conditions, such as infections suffered by patients at the time of blood collection or the administration of steroidal medications that may affect the blood count. Third, we have not studied the correlation between NLR and TME. Fourth, the study does not include patients treated osimertinib as the primary treatment.

In conclusion, our study presents NLR as a significant predictor of clinical efficacy (response and PFS) and OS in NSCLC patients positive for EGFR-mutations and treated with EGFR-TKI including gefitinib and erlotinib. Whether NLR predicts outcome similarly with osimertinib treatment, along with gefitinib and erlotinib, in NSCLC patient’s positive for EGFR-mutations remains to be validated as an immediate follow up to this study.