Association between survivin genetic polymorphisms and epidermal growth factor receptor mutation in non-small-cell lung cancer.

Survivin is an anti-apoptotic protein that is implicated in the regulation of apoptosis and cell cycle in various types of cancers. The current study explored the effect of survivin gene polymorphisms and EGFR mutations in non-small-cell lung carcinoma (NSCLC) patients. A total of 360 participants, including 291 adenocarcinoma lung cancer and 69 squamous cell carcinoma lung cancer patients, were selected for the analysis of three survivin genetic variants (survivin -31, +9194, and +9809) by using real-time PCR genotyping. The results indicated that GC+CC genotypes of survivin -31 were significant association with EGFR mutation in lung adenocarcinoma patients (adjusted odds ratio=3.498, 95% CI = 1.171-10.448; p<0.01). Moreover, The GC+CC genotypes of survivin -31 were associated with EGFR L858R mutation but not in exon 19 in-frame deletions. Furthermore, among patients in exon 19 in-frame deletions, those who have at least one polymorphic G allele of survivin -31 have an increased incidence to develop late-stage when compared with those patients homozygous for C/C (OR, 4.800; 95% CI, 1.305-17.658). In conclusion, our results showed that survivin genetic variants were related to EGFR mutation in lung adenocarcinoma patients and might contribute to pathological development to NSCLC.

Introduction

The lung cancer is the leading cause of death worldwide. In 2009, molecular target therapy such as epidermal growth factor receptor (EGFR) inhibitors was recognized as a potential treatment for certain types of lung cancer 1, 2. The effect of EGFR tyrosine kinase inhibitors (EGFR-TKI) has been linked to EGFR mutations in cancer cells, most of which are exon 19 deletions followed by exon 21 L858R mutations 3-5. Current studies on EGFR mutations have shown a higher tendency of EGFR mutations in adenocarcinoma, such as in female cancer patients with no smoking history 6-8. Regarding racial differences in EGFR mutations, EGFR mutations have been observed in approximately 12%-15% of Caucasian patients with lung adenocarcinoma, but the rate can be as high as 60% in Asian populations 5, 9. For EGFR-TKIs, the most frequently studied drugs are gefitinib and erlotinib 4, 10. Nevertheless, even in patients with EGFR mutations, the treatment efficacy of these drugs is only approximately 60%-70% 11. Therefore, identifying biomarkers that enhance the treatment efficacy of these drugs is crucial.

Survivin, also known as BIRC6, located on human chromosome 17q2, is a member of the inhibitors of apoptosis protein (IAP) family and its key function is apoptosis suppression 12, 13. Researchers have found that in the mitochondria, survivin directly suppresses Bax- and Fas-induced apoptosis and blocks the apoptosis pathway by binding to activated capspase-3 and caspase-7 proteins 14, 15. In addition to apoptosis suppression, an increasing number of studies are showing that survivin is tumor-specific because it is expressed in large quantities in tumor tissues and is closely associated with tumor differentiation, proliferation, and metastasis 16, 17. In non-small-cell lung cancer (NSCLC), high expression of survivin indicates a poor clinical prognosis 18-20. Other studies have suggested that suppressing survivin in lung cancer cells can reduce lung cancer metastasis and invasion 21-23. Some researchers have indicated that in EGFR-mutated lung cancer cell lines, EGFR-TKIs may induce apoptosis by suppressing survivin expression 24-26. A study by Shi showed that survivin expression in the blood is a reliable marker of EGFR-TKI treatment efficacy in patients with lung cancer 26.

Some studies have reported that single nucleotide polymorphism (SNP) of survivin promoters alters protein expression by affecting the functions of transcription factors 27-29. In fact, the SNP of survivin influences the severity and prognosis of many types of cancer including stomach, colorectal, and lung cancer 27, 30-32. Although the SNP of survivin -31 G/C and other SNPs that can affect survivin expression, the association between the SNP of survivin and EGFR mutations in NSCLC still needs to be verified. Moreover, the high EGFR mutation rate, including L858R in exon 21 or in-frame deletion in exon 19, was found in Taiwan populations. Therefore, the present study examined the association between survivin SNP and EGFR mutations and explored the association between survivin SNP and the clinicopathological characteristics in NSCLC.

Methods

Patient Specimens

In 2012-2014, we recruited 360 patients with lung cancer, including 291 adenocarcinoma lung cancer and 69 squamous cell carcinoma lung cancer patients, at Cheng-Ching General Hospital in Taichung, Taiwan. Demographic characteristics and medical information of the patients, including TNM clinical staging, primary tumor size, lymph node involvement, and histologic grade, was obtained from their medical records. Exons 18-21 of the EGFR gene were amplified using polymerase chain reaction and subsequently sequenced as described previously 33. This study was approved by the Institutional Review Board of Cheng-Ching General Hospital (No: HP120009) and informed consent was obtained from all subjects.

Genomic DNA extraction and survivin genotyping

DNA was extracted from buffy coats using a QIAamp DNA blood mini kits (Qiagen, Valencia, California) as described in detail previously 34. DNA was dissolved in TE buffer and used as the template in polymerase chain reactions. Allelic discrimination of survivin -31, +9194, and +9809 gene polymorphism was assessed with the ABI StepOne™ Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) and analyzed using SDS vers. 3.0 software (Applied Biosystems), with the TaqMan assay 28, 29.

Statistical analysis

The distributions of demographic characteristics and genotype frequencies between adenocarcinoma lung cancer and squamous cell carcinoma lung cancer as well as clinicopathological features in different genotypes were analyzed by χ2-test. The odds ratio and 95% CIs of the association between the genotype frequencies and EGFR mutation risk and the clinical pathological characteristics were estimated using multiple logistic regression models after controlling for other covariates. A p value of <0.05 was considered statistically significant. The data were analyzed with SAS statistical software (SAS Institute Inc., Cary, NC, USA).

Results

Patients' characteristics and distribution of lung cancer

Total 360 patients were enrolled in this study. The demographics and clinical characteristics of patients were shown in Table 1. The average age of patients was 66 years. The gender distribution in patients were 205 male (56.9%) and 155 female (44.5%). In all patients, the percentage of adenocarcinoma and squamous cell carcinoma were 80.8% (291/360) and 19.2% (69/360), respectively. Moreover, female patients possessed higher frequency (male vs. female = 49.5% vs. 50.5%) in the adenocarcinoma. For the cigarette smoking status, it was shown 58.6% (205/360) never-smokers and 41.4% (149/360) ever-smokers. Furthermore, never-smoking patients had higher frequency (never-smokers vs. ever-smokers = 66.7% vs. 33.3%) in the adenocarcinoma, while it was shown lower frequency (never-smokers vs. ever-smokers = 24.6% vs. 75.4%) in the squamous cell carcinoma (p<0.001).

Table 1

Demographics and clinical characteristics of 360 patients affected with lung adenocarcinoma and lung squamous cell carcinoma.

Variable	All cases(N=360) n (%)	Adenocarcinoma (N=291) n (%)	Squamous cellcarcinoma (N=69) n (%)	p value
Age
<30	3 (0.8%)	3 (1.0%)	0 (0%)	p=0.963
30-39	5 (1.4%)	4 (1.4%)	1 (1.4%)
40-49	35 (9.7%)	28 (9.6%)	7 (10.1%)
50-59	76 (21.1%)	63 (21.6%)	13 (18.8%)
60-69	78 (21.7%)	61 (21.0%)	17 (24.6%)
≥70	163 (45.3%)	132 (45.4%)	31 (44.9%)
Mean ± SD	66.31 ± 13.89	66.21 ± 13.91	66.71 ± 13.92	p=0.788
Gender
Male	205 (56.9%)	144 (49.5%)	61 (88.4%)	p<0.001
Female	155 (43.1%)	147 (50.5%)	8 (11.6%)
Cigarette smoking status
Never-smoker	211 (58.6%)	194 (66.7%)	17 (24.6%)	p<0.001
Ever-smoker	149 (41.4%)	97 (33.3%)	52 (75.4%)
PPK	48.83 ± 23.84	45.84 ± 21.79	54.42 ± 26.57	p=0.036
Disease stage
IA	35 (9.7%)	33 (11.3%)	2 (2.9%)	p<0.001
IB	45 (12.5%)	40 (13.7%)	5 (7.2%)
IIA	24 (6.7%)	18 (6.2%)	6 (8.7%)
IIB	9 (2.5%)	6 (2.1%)	3 (4.3%)
IIIA	41 (11.4%)	30 (10.3%)	11 (15.9%)
IIIB	58 (16.1%)	33 (11.3%)	25 (36.2%)
IV	148 (41.1%)	131 (45.0%)	17 (24.6%)

Associations between survivin SNPs and lung cancer

The distribution frequency of survivin -31, +9194 and +9809 genotypes in the lung adenocarcinoma and squamous cell carcinoma are shown in Table 2. The alleles with the highest distribution frequency for -31, +9194 and +9809 of survivin in recruited patients with NSCLC were heterozygous C/G, homozygous A/A, and heterozygous T/C, respectively. After adjusting variables, there was no significant difference between the lung adenocarcinoma and squamous cell carcinoma with polymorphisms of the survivin -31, +9194 and +9809 genotypes when compared with wild-type individuals.

Table 2

Distribution frequency of survivin genotypes in 291 lung adenocarcinoma and 69 lung squamous cell carcinoma.

Variable	adenocarcinoma (N=291) (%)	squamous cell carcinoma (N=69) (%)	OR (95% CI)	AOR (95% CI)
survivin-31
CC	91 (31.3%)	22 (31.9%)	1.00	1.00
CG	137 (47.1%)	30 (43.5%)	0.906 (0.492-1.668)	0.776 (0.329-1.828)
GG	63 (21.6%)	17 (24.6%)	1.116 (0.549-2.270)	0.825 (0.298-2.281)
CG+GG	200 (68.7%)	47 (68.1%)	0.972 (0.553-1.708)	0.792 (0.356-1.762)
survivin+9194
AA	172 (59.1%)	43 (62.3%)	1.00	1.00
AG	98 (33.7%)	19 (27.5%)	0.776 (0.428-1.405)	1.291 (0.531-3.138)
GG	21 (7.2%)	7 (10.1%)	1.333 (0.532-3.340)	3.031 (0.839-3.138)
AG+GG	119 (40.9%)	26 (37.7%)	0.874 (0.509-1.500)	1.627 (0.734-3.606)
survivin+9809
TT	108 (37.1%)	25 (36.2%)	1.00	1.00
TC	132 (45.4%)	27 (39.1%)	0.884 (0.485-1.611)	0.864 (0.388-1.923)
CC	51 (17.5%)	17 (24.7%)	1.440 (0.715-2.901)	0.763 (0.281-2.075)
TC+CC	183 (62.9%)	44 (63.8%)	1.039 (0.602-1.792)	0.831 (0.396-1.743)

The AORs with 95% CIs were estimated by multiple logistic regression models after controlling for age, gender and cigarette smoking status.

Patient's characteristics and distribution of EGFR mutations in adenocarcinoma

We further investigated the associations between EGFR mutations and patient's characteristics. As shown in Table 3, both substitution mutation (L858R) and Exon 19 in-frame deletion mutations were shown higher percentage in female patients (male vs. female = 19.2% vs. 80.8% and 44.9% vs. 55.1%, respectively) and in never-smoker patients (never-smokers vs. ever-smokers = 88.5% vs.11.5% and 73.5% vs. 26.5%, respectively). The data of distribution were shown significantly different between control (wild-type) and EGFR mutations in gender (p<0.05) and cigarette smoking status (p<0.05). These results indicated that EGFR mutations were associated with gender and cigarette smoking status.

Table 3

Demographics and clinical characteristics of 190 patients in lung adenocarcinoma with EGFR mutation status.

Variable	Wild type(N=82) n (%)	L858R (N=52) n (%)	In-frame deletion (N=49) n (%)	Others(N=7) n (%)
Age
<30	2 (2.4%)	0 (0%)	1 (2.0%)	0 (0%)
30-39	2 (2.4%)	0 (0%)	1 (2.0%)	0 (0%)
40-49	10 (12.2%)	5 (9.6%)	8 (16.3%)	0 (0%)
50-59	14 (17.1%)	7 (13.5%)	14 (28.6%)	2 (28.6%)
60-69	19 (23.2%)	13 (25.0%)	10 (20.4%)	0 (0%)
≥70	35 (42.7%)	27 (51.9%)	15 (30.6%)	5 (71.4%)
Mean ± SD	65.16 ± 15.06	68.17 ± 12.62	60.98 ± 13.56b	71.57 ± 12.04
Gender
Male	52 (63.4%)	10 (19.2%)a	22 (44.9%) a,b	4 (57.1%)
Female	30 (36.6%)	42 (80.8%)	27 (55.1%)	3 (42.9%)
Cigarette smoking status
Never-smoker	35 (42.7%)	46 (88.5%)a	36 (73.5%) a	4 (57.1%)
Ever-smoker	47 (57.3%)	6 (11.5%)	13 (26.5%)	3 (42.9%)
PPK	51.89 ± 21.46	40.00 ± 26.08	29.00 ± 17.29a	56.67 ± 16.07
Disease stage
IA	8 (9.8%)	4 (7.7%)	5 (10.2%)	1 (14.3%)
IB	6 (7.3%)	8 (15.4%)	7 (14.3%)	1 (14.3%)
IIA	4 (4.9%)	3 (5.8%)	3 (6.1%)	0 (0%)
IIB	0 (0%)	0 (0%)	0 (0%)	0 (0%)
IIIA	11 (13.4%)	7 (13.5%)	2 (4.1%)	0 (0%)
IIIB	14 (17.1%)	5 (9.6%)	3 (6.1%)	1 (14.3%)
IV	39 (47.6%)	25 (48.1%)	29 (59.2%)	4 (57.1%)

aSignificant difference compare with wild type group, p value<0.05.

bSignificant difference compare with L858R group, p value<0.05.

Associations between survivin SNPs and EGFR mutations in adenocarcinoma

To clarify the association between the polymorphism of survivin gene and EGFR mutation, the distribution frequency of survivin (-31, +9194 and +9809) gene genotypes and EGFR mutation type in lung adenocarcinoma patients were estimated. As the results shown in Table 4, GC and GC+CC genotypes of survivin -31 were shown significantly association with EGFR mutation in lung adenocarcinoma patients (AOR = 3.622, 95% CI = 1.158-11.325 and AOR = 3.498, 95% CI = 1.171-10.448, respectively). Moreover, the results of Table 5 were shown that GC+CC genotypes of survivin -31 were shown slightly association with L858R mutation in lung adenocarcinoma patients (AOR = 0.5187, 95% CI = 0.935-2.242). These results indicated the polymorphism of survivin -31 genes were associated with EGFR mutation in adenocarcinoma patients.

Table 4

Distribution frequency of survivin genotypes of 82 EGFR wild type and 108 EGFR mutation type in lung adenocarcinoma patients.

Variable	Wild type (N=82) n (%)	Mutation type (N=108) (%)	OR (95% CI)	AOR (95% CI)
survivin-31
GG	22 (26.8%)	13 (12.0%)	1.00	1.00
GC	39 (47.6%)	58 (53.7%)	2.517 (1.135-5.583)*	3.622 (1.158-11.325)*
CC	21 (25.6%)	37 (34.3%)	2.982 (1.249-7.117)*	3.262 (0.921-11.549)
GC+CC	60 (73.2%)	95 (88.0%)	2.679 (1.256-5.718)*	3.498 (1.171-10.448)*
survivin+9194
AA	51 (62.6%)	54 (50.0%)	1.00	1.00
AG	26 (31.7%)	44 (40.7%)	1.598 (0.862-2.964)	1.363 (0.550-3.376)
GG	5 (6.1%)	10 (9.3%)	1.889 (0.604-5.904)	1.932 (0.348-10.743)
AG+GG	31 (37.8%)	54 (50.0%)	1.645 (0.917-2.951)	1.418 (0.235-8.571)
survivin+9809
TT	30 (36.6%)	43 (39.8%)	1.00	1.00
TC	38 (46.3%)	48 (44.4%)	0.881 (0.469-1.657)	0.561 (0.229-1.376)
CC	14 (17.1%)	17 (15.7%)	0.847 (0.363-1.977)	0.631 (0.170-2.338)
TC+CC	52 (63.4%)	65 (60.2%)	0.872 (0.483-1.576)	1.125 (0.321-3.944)

The AORs with 95% CIs were estimated by multiple logistic regression models after controlling for age and gender.

*p < 0.05.

Table 5

The associations between the polymorphisms of survivin and the EGFR hotspot mutations in lung adenocarcinoma patients.

Variable	Wild type	L858R		Exon 19 in-frame deletion
	(N=82) n (%)	(N=52) (%)	AOR (95% CI)	(N=49) (%)	AOR (95% CI)
survivin-31
GG	22 (26.8%)	6 (11.5%)	1.00	7 (14.3%)	1.00
GC	39 (47.6%)	26 (50.0%)	5.346 (0.906-31.537) p=0.064	27 (55.1%)	2.756 (0.570-13.311) p=0.207
CC	21 (25.6%)	20 (38.5%)	4.948 (0.752-32.575) p=0.096	15 (30.6%)	1.633 (0.245-10.872) p=0.612
GC+CC	60 (73.2%)	46 (88.5%)	5.187 (0.935-2.242) p=0.057	42 (85.7%)	2.460 (0.524-11.539) p=0.254
survivin+9194
AA	51 (62.6%)	26 (50.0%)	1.00	23 (46.9%)	1.00
AG	26 (31.7%)	22 (42.3%)	1.946 (0.495-7.650) p=0.341	21 (42.9%)	2.458 (0.650-9.291) p=0.185
GG	5 (6.1%)	4 (7.7%)	1.241 (0.099-15.603) p=0.867	5 (10.2%)	2.593 (0.205-32.738) p=0.461
AG+GG	31 (37.8%)	26 (50.0%)	0.638 (0.047-8.653) p=0.735	26 (53.1%)	2.481 (0.711-8.650) p=0.154
survivin+9809
TT	30 (36.6%)	16 (30.8%)	1.00	24 (49.0%)	1.00
TC	38 (46.3%)	30 (57.5%)	0.967 (0.265-3.528) p=0.960	16 (32.7%)	0.301 (0.087-1.042) p=0.058
CC	14 (17.1%)	6 (11.5%)	0.312 (0.042-2.308) p=0.254	9 (18.4%)	0.493 (0.080-3.019) p=0.444
TC+CC	52 (63.4%)	36 (69.2%)	0.323 (0.051-2.050) p=0.231	25 (51.0%)	0.339 (0.108-1.063) p=0.064

The AORs with 95% CIs were estimated by multiple logistic regression models after controlling for age and gender.

Associations between survivin SNPs and clinicopathological characteristics of lung cancer

To revealed the association between polymorphisms of survivin gene and different clinical stage of lung cancer in different EGFR mutation of patients. As shown in Table 6, CG+GG genotype of survivin -31 was shown significantly association with clinical advanced stage in lung adenocarcinoma patients with exon 19 mutations (OR = 4.800, 95% CI = 1.305-17.658; p=0.014). These findings indicated that the polymorphisms of survivin -31 may associated with clinical advanced stage of lung cancer.

Table 6

Associations between polymorphic genotypes of survivin -31 and clinicopathologic characteristics of lung cancer.

	Clinical Stage
Variable genotypic frequencies	Stage IA/IB/ IIA/IIB/IIIA/IIIB	Stage IV	OR (95% CI)	p value
All cases (N=360)	(N=212)	(N=148)
survivin -31 CC	75 (35.4%)	38 (25.7%)	1.00
survivin -31 CG+GG	137 (64.6%)	110 (74.3%)	1.585 (0.996-2.520)	p=0.051
adenocarcinoma (N=291)	(N=160)	(N=131)
survivin -31 CC	56 (35.0%)	35 (26.7%)	1.00
survivin -31 CG+GG	104 (56.0%)	96 (73.3%)	1.477 (0.891-2.448)	p=0.129
squamous cell carcinoma (N=69)	(N=52)	(N=17)
survivin -31 CC	19 (36.5%)	3 (17.6%)	1.00
survivin -31 CG+GG	33 (63.5%)	14 (82.4%)	2.687 (0.684-10.561)	p=0.147
Wild type (N=82)	(N=43)	(N=39)
survivin -31 CC	10 (23.3%)	11 (28.2%)	1.00
survivin -31 CG+GG	33 (76.7%)	28 (71.8%)	0.771 (0.286-2.083)	p=0.608
L858R (N=52)	(N=27)	(N=25)
survivin -31 CC	12 (44.4%)	8 (32.0%)	1.00
survivin -31 CG+GG	15 (55.6%)	17 (68.0%)	1.700 (0.548-5.275)	p=0.327
In-frame deletion (N=49)	(N=20)	(N=29)
survivin -31 CC	10 (50.0%)	5 (17.2%)	1.00
survivin -31 CG+GG	10 (50.0%)	24 (82.8%)	4.800 (1.305-17.658)	p=0.014*

*p < 0.05.

Discussion

Survivin is an apoptosis-suppressing protein and its association with many types of cancer (including lung cancer) has been well documented 30, 31, 35, 36. However, few studies have investigated how SNP is related to EGFR mutations or prognosis in lung cancer. The present study revealed that survivin -31 polymorphism may be associated with EGFR mutations, particularly L858R mutations. Furthermore, in patients with exon 19 mutations, survivin -31 polymorphisms have an increased incidence to develop late-stage. To our best knowledge, the present study is the first reports to show a link between survivin SNP and EGFR mutations in patients with lung cancer.

High expression of survivin in the blood and tissues of patients with lung cancer is also associated with clinicopathological characteristics, including metastasis and low survival rates 26, 37-39. Previous studies have shown that high survivin expression in the blood of lung cancer patients undergoing EGFR-TKI treatment is associated with a poor prognosis 26. The present study also identified an association between survivin -31 polymorphism and the EGFR mutation statues, suggesting that EGFR mutations may impair the effect of TKIs; that is, patients with survivin -31G/G genotypes may have higher survivin protein expression. The results were consistent with previous reports that an association between SNP of survivin and lung cancer as well as higher transcriptional activity in the C/C genotype, the latter of which may elevate survivin protein expression 31, 40.

In addition to linking the SNP of survivin -31C/G to the incidence of lung cancer, previous studies have also revealed some clinical manifestations related to survivin -31 C/G polymorphism 36, 41. For example, Javid et al. found that survivin -31C/C in lung cancer patients was associated with a poor overall survival rate 41, and Rosato et al. showed that the -31 C/C genotype was related to node metastasis and patient survival 36. Moreover, Tao et al. showed that survivin 9386 C>T polymorphisms are potential independent prognostic factors in NSCLC patients treated with platinum-based chemotherapy 42. The present study showed that survivin -31C/C was stage-related in patients with EGFR exon 19 mutations. However, this relation was not observed in the EGFR wild type or in lung cancer patients with L858R mutations. A possible explanation is that the effect of survivin differs according to the type of EGFR mutation. For example, Okamoto et al. demonstrated that gefitinib in EGFR-mutated cells modulates cell survival by suppressing survivin expression via the PI3K-AKT signaling pathway 25. Another possibility is that EGFR L858R and exon 19 mutations exists different oncogenic ability 43. Although the previous study showed exon 19 to have a higher oncogenic ability compared with that of L858R 43, the role of survivin in these two distinct mutation processes requires further examination.

In conclusion, our results showed that survivin genetic variants are related to EGFR mutation in lung adenocarcinoma patients and might contribute to pathological development to NSCLC. The findings provide a hint for the genesis of EGFR mutations.