Kinesin family member 2A high expression correlates with advanced tumor stages and worse prognosis in non-small cell lung cancer patients.

BACKGROUND: This present study was to explore the association of kinesin family member 2A (KIF2A) expression with clinicopathological features and survival profiles, and the effect of KIF2A on cell proliferation and chemosensitivity in non-small cell lung cancer (NSCLC).
METHODS: Tumor and paired adjacent specimens were collected from 380 patients with NSCLC underwent resection for immunohistochemistry assay of KIF2A expression. In vitro, the effect of KIF2A on cell proliferation, chemosensitivity to cisplatin/vinorelbine was detected via KIF2A plasmids transfection into NCI-H1299 NSCLC cells.
RESULTS: Kinesin family member 2A expression was upregulated in tumor tissues compared with adjacent tissues, and tumor tissue KIF2A high expression was associated with higher pathological grade (P < .001), larger tumor size (P = .021), lymph node metastasis (P = .044), and increased tumor-node-metastasis stage (P = .001). As for survival profiles, disease-free survival (P < .001) and overall survival (P < .001) were worse in patients with KIF2A high expression compared with those with KIF2A low expression. Multivariate Cox's regression exhibited that KIF2A high expression was an independent predictive factor for lower DFS (P < .001) and OS (P < .001). In vitro, KIF2A promoted proliferation and decreased chemosensitivity to cisplatin but not vinorelbine in NCI-H1299 NSCLC cells.
CONCLUSIONS: The correlation of KIF2A expression with tumor features, survival, and its cellular function implies its potential as a prognostic biomarker and a treatment target in NSCLC.

INTRODUCTION

The global cancer statistics exhibit that lung cancer is one of the leading causes of cancer‐related deaths worldwide with over 1.8 million new cases as well as approximately 1.6 million deaths every year.1 As the main histological type, non‐small cell lung cancer (NSCLC) accounts for approximately 85% of all lung cancer cases.2 Current therapies consist of surgical resection, radiation therapy, chemotherapy regimens (including neoadjuvant/adjuvant therapy), and targeted drugs.3 Although there have been great advancements in the treatment for NSCLC, the 5‐year survival rate of patients with NSCLC is far from satisfaction, which was mainly due to advanced local invasion, distant metastases, tumor relapse, and chemotherapy resistance.4, 5 Thus, it is essential to discover more biomarkers for improving the NSCLC management and prolonging the survival in NSCLC patients.

Kinesin family member 2A (KIF2A), one of the kinesin family members, is identified in mammalian cells and involved in regulating microtubule dynamics during the mitosis via modulating intracellular transport, cell division, and bipolar spindle assembly.6 Previous evidence exhibits that KIF2A serves as a minus‐end depolymerizing motor of the microtubule, which is essential for invading neighboring tissues as well as the lymph node invasion and distant metastasis.7 Furthermore, the carcinogenic effect of KIF2A in development and progression of several cancers, including gastric, breast, and ovarian cancer, has been observed.8, 9, 10 However, for the role of KIF2A in NSCLC, limited information is available.

Thus, we conducted this present study to explore the association of KIF2A expression with clinicopathological features and survival profiles, and the effect of KIF2A on cell proliferation and chemosensitivity in NSCLC.

METHODS

Patients and specimens

In this retrospective study, 380 patients with NSCLC who underwent resection in our hospital between January 2012 and December 2014 were reviewed. The inclusion criteria were as follows: (a) newly diagnosed as primary NSCLC confirmed by clinical, histological, and pathological examinations, (b) received surgical resection without neoadjuvant therapy, (c) tumor and paired adjacent tissue specimens were accessible and available for immunohistochemistry (IHC) assay, and (d) with complete medical data. The exclusion criteria were as follows: (a) secondary NSCLC, (b) patients complicated with other malignant tumors, and (c) tumor specimens or clinical data were missing. A total of 380 formalin‐fixed paraffin‐embedded (FFPE) tumor and paired adjacent specimens were collected from pathology department of our hospital after approval, and all FFPE specimens were eligible for IHC assay. The Ethics Committee of Xi'an No.4 Hospital had approved the current study, and the written informed consents or verbal informed consents with tape recording were obtained from all patients or their guardians.

Data collection

Basic characteristics of each patient were collected from medical records including age, gender, smoke, and drink status, and the tumor‐related features were also extracted including pathological grade, tumor size, lymph node (LYN) metastasis, tumor‐node‐metastasis (TNM) stage (based on the Union for International Cancer Control system [7th edition]), and carcinoembryonic antigen (CEA) level was measured at diagnosis.

KIF2A detection by IHC

The IHC assay for detection of KIF2A expression in the tumor and paired adjacent tissue was performed as follows: FFPE specimens were sliced into 4‐µm sections, and then the sections were deparaffinized by xylene, rehydrated by graded ethanol, and followed by antigen retrieval in citrate buffer. After blocking endogenous peroxidase using 3% H2O2, sections were immersed with goat serum for blocking nonspecific endogenous antigens. Next, the sections were incubated with rabbit polyclonal to KIF2A (Abcam) at 4°C overnight. Negative controls were incubated with phosphate‐buffered saline only. After washing, the sections were incubated with horseradish peroxidase (HRP)‐conjugated goat‐anti‐rabbit IgG antibody (Abcam) at room temperature for 30 minutes. After that, the tissue sections were stained with diaminobenzidine (DAB), counterstained with hematoxylin, and sealed with neutral resin. Finally, the sections were viewed under the light microscope and photographed on a Nikon ECLIPSE E600 microscope (Nikon Instruments). A semi‐quantitative method based on the average intensity and density of positively stained tumor cells was used to assess the KIF2A expression in the tumor and paired adjacent tissue.11 The intensity of positively stained tumor cells was scored as follows: 0 (no staining), 1 (weak staining), 2 (moderate staining), and 3 (strong staining). And the density was assessed by the percentage of positively stained tumor cells, which was scored as follows: 0, 0%; 1, <25%; 2, 26 ~ 50%; 3, 51 ~ 75%; and 4, >75%. The final semi‐quantitative score was calculated by multiplying of intensity score and density score, ranging from 0 to 12. And the final semi‐quantitative score >3 was defined as KIF2A high expression, while the final semi‐quantitative score ≤3 was defined as KIF2A low expression.

Treatment and follow‐up

Appropriate postoperative adjuvant treatment was administered to patients by attending surgeons after considering the TNM stage, surgical margin status, and clinical features. Briefly, according to the surgical margin status, patients with IA stage were closely monitored or treated with reresection or radiotherapy (RT); patients with IB stage were observed or administered with chemotherapy (for high‐risk patients) or reresection ± chemotherapy; patients with IIA ~ IIB stage were given chemotherapy or reresection + chemotherapy, or chemoradiation + chemotherapy; and patients with IIIA stage were treated with chemotherapy + RT, or chemoradiation + chemotherapy. And the common chemotherapy regimens were cisplatin‐based, including cisplatin + vinorelbine, cisplatin + etoposide, cisplatin + vinblastine, cisplatin + gemcitabine, and cisplatin + docetaxel. The adjuvant chemotherapy was repeated every 21 days for a total of 4 cycles, which was initiated at 6 ~ 8 weeks postoperation, and the RT was started at 120 ~ 140 days postoperation. Besides, patients were followed up by clinic visit or telephone calls regularly. Patients’ follow‐up records were reviewed, and the survival data were collected. The disease‐free survival (DFS) and overall survival (OS) were evaluated with the last follow‐up at December 31, 2018.

Cell culture

The human NSCLC cell line NCI‐H1299 was purchased from American Typical Culture Collection (ATCC) and cultured in Roswell Park Memorial Institute 1640 Medium (ATCC) supplemented with 10% fetal bovine serum (Life Technologies). Incubation of cells was carried out in a 5% CO2 humid atmosphere at 37°C.

Transfection

The over‐expression plasmids and the knockdown plasmids were constructed by Shanghai QeeJen Bio‐Tech Co., Ltd. The pEX2 was used to construct the over‐expression plasmids, and the pGPU6 was used to construct knockdown plasmids. The constructed plasmids were transfected into NCI‐H1299 cells using HilyMax (Dojindo). The NCI‐H1299 cells were divided into four groups: KIF2A over‐expression group (KIF2A (+)) that was transfected with KIF2A over‐expression plasmids; over‐expression negative control group (NC (+)) that was transfected with NC over‐expression plasmids; KIF2A knockdown group (KIF2A (−)) that was transfected with KIF2A knockdown plasmids; and knockdown negative control group (NC (−)) that was transfected with NC knockdown plasmids.

Cell proliferation detection

At 0, 24, 48, and 72 hours after transfection, the cell proliferation was measured by the CCK‐8 (Dojindo) assay according to the manufacturer's instructions and calculated by optical density (OD) value.

Drug sensitivity detection

Cisplatin and vinorelbine were purchased from Sigma‐Aldrich (Sigma‐Aldrich). Cisplatin was dissolved in phosphate buffer saline to obtain store solutions with following concentrations: 1, 2, 4, 8, 16, and 32 μmol/L. And vinorelbine was dissolved in DMSO to obtain store solutions with concentrations of 0.2, 0.4, 0.8, 1.6, 3.2, and 6.4 μmol/L. At 24 hours after transfection, cells in four groups were plated in 96‐well plate and cultured with cisplatin or vinorelbine store solutions for additional 24 hours, respectively. Then, the cell viability in four groups was detected by CCK‐8 (Dojindo) assay, and the relative cell viability (%) was calculated by setting corresponding untreated controls (0 μmol/L drug) as 100%. And 50% inhibitory concentration (IC50) of two drugs in four groups was, respectively, calculated.

Statistical analysis

Statistical analyses were performed with the use of SPSS 21.0 (IBM), and figures were plotted using GraphPad Prism 7.00 (GraphPad Software). Data were presented as mean and standard deviation (SD) or count (percentage). The comparison was determined by McNemar's test (for paired comparison), chi‐square test, Wilcoxon's rank‐sum test, or unpaired t test. DFS was calculated from the date of resection to the date of relapse or death, and patients not known to have relapsed or died at last follow‐up were censored on the date they were last examined. OS was calculated from the date of resection to the date of death, and patients not known to have died at last follow‐up were censored on the date they were last known to be alive. Also, patients who lost follow‐up were censored on the date that they were last known to be alive. DFS and OS were displayed by Kaplan‐Meier curve, and the difference of DFS or OS between groups was determined by log‐rank test. The factors predicting DFS or OS were analyzed by univariate and multivariate Cox's proportional hazard regression model. Drug concentration required to inhibit growth by 50% (IC50) was calculated by probit regression model. P value <.05 was considered significant.

RESULTS

Overview characteristics of NSCLC patients

A total of 380 patients with NSCLC were enrolled in the present study. In total patients, there were 182 (47.9%) patients younger than 60 years old and 198 (52.1%) patients older than 60 years old. And among all the patients with NSCLC, there were 87 (22.9%) females and 293 (77.1%) males. As for pathological grade, there were 56 (14.7%), 237 (62.4%), and 87 (22.9%) patients with pathological grade G1, G2, and G3, respectively. Regarding TNM stage, there were 115 (30.3%), 133 (35.0%) and 132 (34.7%) patients with TNM stage I, II, and III, respectively.

Expression of KIF2A in tumor tissues and adjacent tissues

Expression of KIF2A in tumor tissues and adjacent tissues were evaluated by IHC (Figure 1). There were 199 (52.4%) tumor tissues with KIF2A high expression and 181 (47.6%) tumor tissues with KIF2A low expression; there were 148 (38.9%) adjacent tissues with KIF2A high expression and 232 (61.1%) adjacent tissues with KIF2A low expression (Table 1). Further analysis indicated that KIF2A expression was upregulated in tumor tissues compared with adjacent tissues (P < .001).

Figure 1

Expression of KIF2A in NSCLC tumor tissues and adjacent tissues. Examples of IHC images about KIF2A high expression in tumor tissues (*200; *100) and KIF2A low expression in adjacent tissues (*200; * 100). IHC, immunohistochemistry; KIF2A, kinesin family member 2A; NSCLC, non‐small cell lung cancer

Table 1

Comparison of KIF2A expression in tumor tissues and adjacent tissues

Items	KIF2A expression		P value
	High	Low
Tumor tissues, No. (%)	199 (52.4)	181 (47.6)	<.001
Adjacent tissues, No. (%)	148 (38.9)	232 (61.1)

Comparison was determined by McNemar's test.

Abbreviation: KIF2A, kinesin superfamily protein 2A.

Correlation of KIF2A expression with tumor features in NSCLC patients

All patients with NSCLC were divided into KIF2A high expression and KIF2A low expression according to the cutoff value of KIF2A IHC score of tumor tissue at baseline (Table 2). KIF2A high expression was associated with higher pathological grade (P < .001), larger tumor size (P = .021), LYN metastasis (P = .044), higher TNM stage (P = .001), but not with age (P = .949), gender (P = .109), smoke (P = .364), drink (P = .465), and CEA (P = .055) in patients with NSCLC.

Table 2

Comparison of clinical features between KIF2A high and low expression patients

Items	Total patients (N = 380)	KIF2A expression		P value
		High (n = 199)	Low (n = 181)
Age, No. (%)
≤60 y	182 (47.9)	95 (47.7)	87 (48.1)	.949
>60 y	198 (52.1)	104 (52.3)	94 (51.9)
Gender, No. (%)
Female	87 (22.9)	39 (19.6)	48 (26.5)	.109
Male	293 (77.1)	160 (80.4)	133 (73.5)
Smoke, No. (%)
No	173 (45.5)	95 (47.7)	78 (43.1)	.364
Yes	207 (54.5)	104 (52.3)	103 (56.9)
Drink, No. (%)
No	234 (61.6)	126 (63.3)	108 (59.7)	.465
Yes	146 (38.4)	73 (36.7)	73 (40.3)
Pathological gradea, No. (%)
G1	56 (14.7)	20 (10.0)	36 (19.9)	<.001
G2	237 (62.4)	119 (59.8)	118 (65.2)
G3	87 (22.9)	60 (30.2)	27 (14.9)
Tumor size, No. (%)
≤5 cm	216 (56.8)	102 (51.3)	114 (63.0)	.021
>5 cm	164 (43.2)	97 (48.7)	67 (37.0)
LYN metastasis, No. (%)
No	247 (65.0)	120 (60.3)	127 (70.2)	.044
Yes	133 (35.0)	79 (39.7)	54 (29.8)
TNM stage, No. (%)
I	115 (30.3)	45 (22.6)	70 (38.7)	.001
II	133 (35.0)	74 (37.2)	59 (32.6)
III	132 (34.7)	80 (40.2)	52 (28.7)
CEA, No. (%)
≤5 ng/mL	155 (40.8)	72 (36.2)	83 (45.9)	.055
>5 ng/mL	225 (59.2)	127 (63.8)	98 (54.1)

Comparison was determined by chi‐square test or Wilcoxon's rank‐sum test.

Abbreviations: CEA, carcinoembryonic antigen; KIF2A, kinesin superfamily protein 2A; LYN, lymph node.

G1, well differentiation; G2, moderate differentiation; and G3, poor differentiation.

Correlation of KIF2A expression with survival in NSCLC patients

DFS was reduced in patients with KIF2A high expression compared with those with KIF2A low expression (P < .001) (Figure 2A). In addition, OS was also decreased in patients with KIF2A high expression compared with those with KIF2A low expression (P < .001) (Figure 2B).

Figure 2

Correlation of KIF2A expression with DFS and OS. Comparison of DFS between patients with KIF2A high expression and low expression (A). Comparison of OS between patients with KIF2A high expression and low expression (B). The survivals for NSCLC patients were exhibited by Kaplan‐Meier curve, and the comparison of survival between patients with KIF2A high expression and low expression was performed by log‐rank test. P < .05 was considered significant. DFS, disease‐free survival; KIF2A, kinesin family member 2A; NSCLC, non‐small cell lung cancer; OS, overall survival

Factors affecting DFS in NSCLC patients

Univariate Cox's regression analysis presented that KIF2A high expression (P < .001), higher pathological grade (P < .001), larger tumor size (>5 cm) (P < .001), LYN metastasis (P < .001), higher TNM stage (P < .001), and higher CEA level (>5 ng/mL) (P = .018) were associated with lower DFS in patients with NSCLC (Table 3). Multivariate Cox's regression exhibited that KIF2A high expression (P < .001), higher pathological grade (P = .003), LYN metastasis (P < .001), and higher CEA level (5 ng/mL) (P = .024) were independent predictive factors for worse DFS in NSCLC patients.

Table 3

Univariate and multivariate Cox's proportional hazard regression model analyses of factors predicting DFS

Items	Univariate Cox's regression		Multivariate Cox's regression
	P value	HR (95%CI)	P value	HR (95%CI)
KIF2A high expression	<.001	1.725 (1.368‐2.173)	<.001	1.587 (1.233‐2.042)
Age (>60 y)	.124	1.196 (0.952‐1.503)	.431	1.100 (0.868‐1.393)
Male	.638	1.069 (0.811‐1.409)	.325	0.857 (0.630‐1.165)
Smoke	.289	1.132 (0.901‐1.422)	.352	1.122 (0.880‐1.432)
Drink	.173	1.176 (0.932‐1.483)	.067	1.247 (0.985‐1.579)
Higher pathological grade	<.001	1.537 (1.281‐1.845)	.003	1.355 (1.110‐1.654)
Tumor size (>5 cm)	<.001	1.595 (1.269‐2.005)	.228	0.819 (0.592‐1.133)
LYN metastasis	<.001	2.675 (2.109‐3.393)	<.001	2.401 (1.779‐3.242)
Higher TNM stage	<.001	1.537 (1.331‐1.776)	.132	1.178 (0.952‐1.457)
CEA (>5 ng/mL)	.018	1.322 (1.049‐1.668)	.024	1.316 (1.037‐1.671)

Abbreviations: CEA, carcinoembryonic antigen; CI, confidence; DFS, disease‐free survival; HR, hazard ratio; KIF2A, kinesin superfamily protein 2A; LYN, lymph node.

Factors affecting OS in NSCLC patients

Univariate Cox's regression revealed that KIF2A high expression (P < .001), higher pathological grade (P < .001), larger tumor size (>5 cm) (P < .001), LYN metastasis (P < .001), higher TNM stage (P < .001), and higher CEA level (>5 ng/mL) were correlated with decreased OS in patients with NSCLC (Table 4). Multivariate Cox's regression showed that KIF2A high expression (P < .001), higher pathological grade (P = .002), LYN metastasis (P < .001), and higher CEA level (>5 ng/mL) (P < .002) were independent predictive factors for reduced OS in patients with NSCLC.

Table 4

Univariate and multivariate Cox's proportional hazard regression model analyses of factors predicting OS

Items	Univariate Cox's regression		Multivariate Cox's regression
	P value	HR (95%CI)	P value	HR (95%CI)
KIF2A high expression	<.001	1.779 (1.378‐2.296)	<.001	1.708 (1.294‐2.254)
Age (>60 y)	.697	1.051 (0.819‐1.348)	.513	0.917 (0.708‐1.188)
Male	.650	0.934 (0.696‐1.253)	.062	0.731 (0.526‐1.015)
Smoke	.454	1.100 (0.857‐1.412)	.504	1.095 (0.838‐1.431)
Drink	.282	1.149 (0.892‐1.481)	.100	1.243 (0.959‐1.612)
Higher pathological grade	<.001	1.652 (1.355‐2.015)	.002	1.413 (1.135‐1.760)
Tumor size (>5 cm)	<.001	1.888 (1.471‐2.422)	.575	1.108 (0.775‐1.585)
LYN metastasis	<.001	3.419 (2.647‐4.416)	<.001	3.295 (2.373‐4.574)
Higher TNM stage	<.001	1.545 (1.319‐1.810)	.652	0.946 (0.741‐1.206)
CEA (>5 ng/mL)	.002	1.516 (1.171‐1.961)	.002	1.531 (1.171‐2.000)

Abbreviations: CEA, carcinoembryonic antigen; CI, confidence; HR, hazard ratio; KIF2A, kinesin superfamily protein 2A; LYN, lymph node; OS, overall survival.

Effect of KIF2A on cell proliferation and chemosensitivity in NSCLC cells

In NCI‐H1299 NSCLC cells, cell proliferation was increased in KIF2A (+) group compared with NC (+) group at 48 hours (P < .05) and 72 hours (P < .05) after transfection, while reduced in KIF2A (−) group compared with NC (−) group at 72 hours (P < .05) (Figure 3A). As for cisplatin sensitivity, relative cell viability was increased in KIF2A (+) group compared with NC (+) group in 2 μmol/L (P < .01), 4 μmol/L (P < .05), 8 μmol/L (P < .05), and 16 μmol/L (P < .05) cisplatin‐treated cells, whereas it was reduced in KIF2A (−) group compared with NC (−) group in 2 μmol/L (P < .05), 4 μmol/L (P < .05), 8 μmol/L (P < .05), and 16 μmol/L (P < .01) cisplatin‐treated cells (Figure 3B). However, as for vinorelbine sensitivity, relative cell viability was similar between KIF2A (+) group and NC (+) group, between KIF2A (−) group and NS (−) group in 0.2, 0.4, 0.8, 1.6, 3.2, and 6.4 μmol/L vinorelbine‐treated cells (Figure 3C). Meanwhile, the IC50 value of cisplatin was increased in KIF2A (+) group compared with NC (+) group (P < .05), while decreased in KIF2A (−) group compared with NC (−) group, whereas IC50 value of vinorelbine was of no difference between KIF2A (+) group and NC (+) group, or between KIF2A (−) group and NC (−) group (both P > .05) (Figure 3D).

Figure 3

Effect of KIF2A on cell proliferation and chemosensitivity. Comparison of cell proliferation between NC(+) and KIF2A(+), between NC(−) and KIF2A(−) (A). Comparison of relative cell viability between NC(+) and KIF2A(+), between NC(−) and KIF2A(−) in 1, 2, 4, 8, 16, and 32 μmol/L cisplatin‐treated cells (B). Comparison of relative cell viability between NC(+) and KIF2A(+), between NC(−) and KIF2A(−) in 0.2, 0.4, 0.8, 1.6, 3.2, and 6.4 μmol/L vinorelbine‐treated cells (C). Comparison of IC50 of cisplatin and vinorelbine between NC(+) and KIF2A(+) as well as between NC(−) and KIF2A(−) (D). Comparison was determined by the unpaired t test. IC50 was calculated by probit regression model. P < .05 was considered significant. *P < .05, **P < .01. IC50, 50% inhibitory concentration; KIF2A(+), KIF2A over‐expression group; NC(+), over‐expression negative control group; KIF2A(−), KIF2A knockdown group; and NC(−), knockdown negative control group

DISCUSSION

In the present study, we found that (a) KIF2A expression was upregulated in tumor tissues compared with adjacent tissues, and its high expression was associated with increased pathological grade, larger tumor size, LYN metastasis, and advanced TNM stage in patients with NSCLC. (b) KIF2A high expression was an independent predictive factor for decreased DFS and OS in patients with NSCLC. (c) KIF2A promoted cell proliferation and repressed cell chemosensitivity to cisplatin in NSCLC.

The kinesin proteins are microtubules depolymerases, which are involved in mitosis, cell migration, trafficking, and cell signaling.12, 13 KIF2A, as one member of the kinesin protein family, has attracted a lot of attention due to its implication in the development and progression of several tumors.8, 9, 10 For example, in gastric cancer, KIF2A is upregulated in cancerous tissues compared with adjacent non‐tumor tissues and its high expression is correlated with advanced TNM stage and LYN metastasis.9 In another study, the expression of KIF2A is increased in tumor tissues compared with normal adjacent tissues from patients with breast cancer and positively associated with LYN metastasis and HER2 status.14 These studies suggest the close association of KIF2A with unfavorable clinical features in several cancers. As for in lung cancer, one previous study with small sample (77 cases) indicates that KIF2A is upregulated in lung adenocarcinoma tissues compared with adjacent normal tissues and correlates with TNM stage and lymph node metastasis; in addition, its hyper‐expression in lung adenocarcinoma is an independent predictive factor for worse OS in patients with lung adenocarcinoma.15 However, this previous study has a relatively small sample size, which may lead to reduced validation. Notably, DFS, which serves as a key index for assessing prognosis, and the drug sensitivity of KIF2A are not detected in the previous study. In order to further validate the role of KIF2A in development and progression of NSCLC, we conducted the present study with 380 patients with NSCLC to explore the association of KIF2A expression with DFS and OS, and the effect of KIF2A on cell proliferation and chemosensitivity. We found that KIF2A was upregulated in tumor tissues compared with adjacent tissues, and its high expression was associated with higher pathological grade, larger tumor size, LYN metastasis, and advanced TNM stage in patients with NSCLC. The possible reasons might include: (a) High expression of KIF2A might activate its downstream oncogenic signaling pathways, such as PI3K/AKT signaling pathway, which might promote the development of NSCLC.16 Therefore, KIF2A was upregulated in tumor tissues compared with adjacent tissues. (b) According to the previous study, KIF2A serves as a microtubule‐associated protein and increase cell mitosis, which leads to the enhanced microtubules turnover and increased cell motility in NSCLC. Thus, high expression of KIF2A was associated with advanced tumor features in patients with NSCLC.17

The predictive value of KIF2A in prognosis is reported in several tumors.8, 9, 10, 14 One study exhibits a correlation between KIF2A expression and the 5‐year survival rate in patients with gastric cancer.9 Another study suggests that KIF2A expression is an independent predictive factor for worse survival in patients with breast cancer.14 Similarly, in patients with NSCLC, we also observed that KIF2A was associated with decreased DFS and OS; meanwhile, KIF2A high expression was an independent predictive factor for reduced DFS and OS. The possible reasons might be that (a) KIF2A high expression was associated with higher pathological grade, larger tumor size, LYN metastasis, and advanced TNM stage. Thus, KIF2A might indirectly affect prognosis via interaction with these clinicopathological properties. (b) KIF2A upregulation decreased chemosensitivity of NSCLC (which was validated in our subsequent cellular experiments), contributing to poor survival in patients with NSCLC.

Existing studies reveal that KIF2A is involved in the regulation of tumor cell activities in several cancers, including ovarian cancer, glioma, and gastric cancer.8, 16, 18 For example, downregulation of KIF2A inhibits ovarian cancer cell proliferation, migration and invasion, but induces apoptosis.8 Another cellular experiment demonstrates that KIF2A silencing suppresses the proliferation, migration, and invasion in glioma cells.18 In ovarian cancer, KIF2A shows a similar effect on cell proliferation, migration, and invasion, which is suppressed by its target gene miR‐206.8 As for the underlying mechanism of KIF2A in lung cancer, only one previous study indicates that KIF2A knockdown inhibits cancer cell proliferation and motility but induces apoptosis in lung squamous cell carcinoma cells.19 In the present study, we further validated the effect of KIF2A on cell proliferation and chemosensitivity via transfection of both KIF2A over‐expression plasmids and knockdown plasmids in NSCLC cells, and we observed that KIF2A over‐expression increased cell proliferation and decreased chemosensitivity to cisplatin but not vinorelbine, and KIF2A knockdown decreased cell proliferation and enhanced chemosensitivity to cisplatin but not vinorelbine in NSCLC cells. The possible reasons might include as follows: (a) KIF2A might activate oncogenic signaling pathway, such as MAPK pathway, which led to increased cell proliferation in NSCLC.20 (b) According to the previous study, cisplatin served as a well‐known chemotherapeutic drug via interfering with DNA repair mechanisms, causing DNA damage and subsequently inducing apoptosis in cancer cancers, and the effect of cisplatin could be inhibited through regulating mitochondrial fission.21, 22 Furthermore, KIF2A regulates microtubule dynamics and enhances the mitosis, which affects mitochondrial fission. Therefore, KIF2A might decrease the cell chemosensitivity to cisplatin.23 (c) As for vinorelbine, it exerts anti‐tumor properties and induces cell death via enhancing inflammation and blocking microtubules.24 Therefore, vinorelbine might inhibit the effect of KIF2A on microtubule regulation during the mitosis, and KIF2A might have no influence of cell chemosensitivity to vinorelbine in NSCLC treatment. These experimental data further explained our clinical findings that KIF2A was correlated with advanced tumor features and worse survival profile in NSCLC.

In summary, KIF2A is upregulated in tumor tissues and correlates with advanced tumor features as well as unfavorable survival profiles in patients with NSCLC, and it promotes proliferation while decreases chemosensitivity to cisplatin in NSCLC cells.