CDK14 expression is elevated in patients with non-small cell lung cancer and correlated with poor prognosis.

OBJECTIVE: To investigate the clinical significance of cyclin-dependent kinase 14 (CDK14) expression in patients with non-small cell lung cancer (NSCLC).
METHODS: The present prospective observational study included 193 patients diagnosed with NSCLC between January 2010 and December 2014. NSCLC tumor tissues and paired paracancerous normal tissues were obtained from all patients. CDK14, thyroid transcription factor 1 (TTF-1), cytokeratin 5/6 (CK5/6), and Ki67 expression was measured via immunohistochemistry (IHC).
RESULTS: CDK14 staining was strong (>3) in 129 patients (66.49%) and weak (≤3) in 64 patients (33.16%). The mean IHC scores were markedly higher in tumor tissues than in paracancerous tissues. Pearson's analysis demonstrated that the IHC scores of CDK14 expression were positively correlated with TTF-1, CK5/6, and Ki67 scores. Kaplan-Meier analysis illustrated that 5-year overall survival was markedly longer in patients with weak CDK14 staining. TNM stage, pleural invasion, lymph node metastasis, CDK14 expression, and Ki67 expression were risk factors for 5-year overall survival in patients with NSCLC.
CONCLUSION: CDK14 overexpression portended poor outcomes in patients with NSCLC, and CDK14 expression was correlated with TTF-1, CK5/5, and Ki67 expression.

Introduction

Non-small cell lung cancer (NSCLC) is the most common lung cancer, accounting for almost 85% of lung cancer cases.[1,2] Currently, several cancer biomarkers are used to diagnose NSCLC, such as thyroid transcription factor 1 (TTF-1), napsin A, cytokeratin 7 (CK7), CK5/6, and Ki67.[3-5] However, because of its mild symptoms, NSCLC is generally not diagnosed in early stages, leading to a poor prognosis and low 5-year survival rate.

Cyclin-dependent kinase 14 (CDK14) is a member of CDK family. Currently, most CDKs are considered cancer promoters, including CDK2, CDK4, CDK6, CDK9, CDK12, CDK14, and CDK16.[7-9] In lung cancer, several studies also found that CDK family members facilitate oncogenesis and cancer development. Wang et al. demonstrated that CDK inhibitors could be used to treat solid tumors, including lung cancer, by regulating the cell cycle and gene transcription. Danilov et al. revealed that dinaciclib could induce anaphase catastrophe in lung cancer cells by suppressing CDK1 and CDK2. Other studies also found that inhibitors of CDKs, such as CDK4/6 dual inhibitors, have the potential to be used in lung cancer treatment.[12,13] However, few studies focused on the role of CDK14 in lung cancer.

We conducted observational research to investigate the clinical significance of CDK14 in patients with NSCLC. We observed that patients with strong CDK14 expression had worse prognosis and shorter 5-year overall survival (OS). This research might provide additional clinical evidence of the role of CDK14 in NSCLC.

Materials and methods

Subjects and tissue samples

The present prospective observational study included patients diagnosed with NSCLC for the first time as confirmed via histological analysis between January 2010 and December 2014. No patient had received chemotherapy or radiotherapy prior to enrollment. Patients’ pathological types were identified. The exclusion criteria were diagnoses of other primary cancers, cancer metastasis to the lungs, the presence of severe infections such as pneumonia or pulmonary tuberculosis, and the presence of severe renal, liver, or cardiovascular diseases. All patients provided written informed consent. The present study was approved by the Ethic Committee of Qujing No. 1 Hospital (approval number: QJH-2009-4-13).

Immunohistochemistry (IHC) for CDK14

NSCLC tumor tissues and paired paracancerous normal tissues were obtained from all patients. For IHC analysis, the tissues were fixed, embedded, sectioned, and stained with hematoxylin and eosin. After immersion in 3% H2O2, samples were incubated with anti-CDK14 primary antibody (PFTK1, Abcam, Cambridge, UK) overnight at 4°C, followed by incubation with Goat Anti-Rabbit IgG H&L (HRP) (ab205718, Abcam) at 37°C for 45 minutes. Samples were then stained with diaminobenzidine, and staining was quantified using the Allred scoring system. The total IHC score was calculated as the sum of the scores for staining intensity (0, no staining; 1, weak staining; 2, moderate staining; 3, strong staining) and the percent area of staining (0, 0%–10%; 1, 11%–25%; 2, 26%–50%; 3, 51%–75%; 4, 76%–100%). IHC scores of >3 were defined as strong staining and high expression, and scores of ≤3 were defined as weak staining and low expression.

Data collection and follow-up

Patient characteristics, including age, sex, body mass index (BMI), tumor-nodes-metastasis (TNM) stage, pathological types, and smoking habits, were recorded. TTF-1, CK5/6, and Ki67 expression in tumor tissues was measured using IHC. All patients were followed up for 5 years. For survival analysis, OS was defined as the time from admission to death or the last follow-up.

Statistical analysis

Data were presented as the mean ± standard deviation. The chi-square test was used to analyze odds ratios. Student’s t-test was used for comparisons between two groups. Pearson’s correlation analysis was used to assess correlations between factors. Kaplan–Meier curves were generated for survival analysis. P < 0.05 indicated statistical significance. Logistic regression was performed to identify risk factors for mortality. All statistical analyses were performed using SPSS 20.0 (IBM Corp., Armonk, NY, USA).

Results

Patient characteristics

As presented in Table 1, the study cohort included 193 patients (119 men, 74 women) with a mean age of 53.01 ± 11.56 years. The mean BMI of the patients was 22.75 ± 3.02 kg/m2. Seventy-nine patients (40.93%) had TNM stage I or II lesions, whereas 114 patients (59.07%) had TNM stage III or IV lesions. Squamous cell carcinoma was found in 106 patients (54.92%), and adenocarcinoma was found in 87 patients (45.08%).

Table 1.

Characteristics of all patients with non-small cell lung cancer.

Variables
Age, years	53.01 ± 11.56
BMI, kg/m2	22.75 ± 3.02
Sex, male:female	119:74
Current smoker, n (%)	107 (55.44)
TNM stage, n (%)
I–II	79 (40.93)
III–IV	114 (59.07)
Pathological type, n (%)
Squamous cell carcinoma	106 (54.92)
Adenocarcinoma	87 (45.08)
Lymph node metastasis, n (%)	128 (65.98)
Pleural invasion, n (%)	103 (53.57)
CDK14 staining, n (%)
Strong	129 (66.49)
Weak	64 (33.16)

BMI, body mass index; CDK14, cyclin-dependent kinase 14.

CDK14 expression was elevated in NSCLC tumor tissues

The IHC results illustrated that CDK14 expression was strong in 129 patients (66.49%) and weak in 64 patients (33.16%, Table 1). Meanwhile, the mean IHC score was markedly higher in tumor tissues than in paracancerous normal tissues (P < 0.05, Figure 1). No significant difference in the mean IHC score was detected between squamous cell carcinoma and adenocarcinoma lesions. This result suggested that CDK14 expression was elevated in NSCLC tumors.

Figure 1.

CDK14 expression in tumor and paracancerous normal tissues from patients with NSCLC. (a) CDK14 expression in NSCLC tumor and paracancerous normal tissues in patients with squamous cell carcinoma and adenocarcinoma. (b) Mean immunohistochemistry scores in different tissue samples from patients with squamous cell carcinoma and adenocarcinoma.

CDK14, cyclin-dependent kinase 14; NSCLC, non-small cell lung cancer.

Correlations among CDK14, TTF-1, CK5/6, and Ki67 expression

We next investigated the relationships among the NSCLC biomarkers TTF-1, CK5/6, Ki67, and CDK14. Patients with strong CDK14 staining were significantly more likely to have strong TTF-1, CK5/6, and Ki67 staining than those with weak CDK14 staining (all P < 0.05, Table 2). The mean IHC score for TTF-1 was significantly higher in patients with squamous cell carcinoma than in those with adenocarcinoma, whereas the score for CK5/6 was markedly higher in patients with adenocarcinoma (both P < 0.05). Pearson’s correlation analysis demonstrated that the IHC score for CDK14 expression was positively correlated with those for TTF-1, CK5/6, and Ki67 (all P < 0.05, Table 3).

Table 2.

Correlations of CDK14 staining with TTF-1, CK5/6, and Ki67 staining in patients with non-small cell lung cancer.

Variables, n (%)	Strong CDK14 staining (n = 129)	Weak CDK14 staining (n = 64)	P
TTF-1			<0.001
Strong	113 (87.60)	13 (20.31)
Weak	16 (12.40)	51 (79.69)
CK5/6			<0.001
Strong	91 (70.54)	22 (34.38)
Weak	38 (29.46)	42 (65.63)
Ki67			<0.001
Strong	86 (66.67)	20 (31.25)
Weak	43 (33.33)	44 (68.75)

CDK14, cyclin-dependent kinase 14; TTF-1, thyroid transcription factor 1; CK5/6, cytokeratin 5/6.

Table 3.

Pearson’s correlation analysis of CDK14, TTF-1, CK5/6, and Ki67 expression in patients with non-small cell lung cancer.

	CDK14	TTF-1	CK5/6	Ki67
CDK14
Pearson’s correlation coefficient	1	0.431	0.255	0.190
P	–	<0.001	<0.001	<0.001
TTF-1
Pearson’s correlation coefficient	0.431	1	0.203	−0.005
P	<0.001	–	0.005	0.943
CK5/6
Pearson’s correlation coefficient	0.255	0.203	1	0.150
P	<0.001	0.005	–	0.037
Ki67
Pearson’s correlation coefficient	0.190	−0.005	0.150	1
P	<0.001	0.943	0.037	-

CDK14, cytokeratin-dependent kinase 14; TTF-1, thyroid transcription factor 1; CK5/6, cytokeratin 5/6.

High CDK14 expression was associated with the clinical characteristics of patients with NSCLC

To further reveal the clinical significance of CDK14 in NSCLC, the relationships between CDK14 expression and clinical characteristics were analyzed. As presented in Table 4, patients with strong CDK14 staining were more likely to have TNM stage III–IV tumors, pleural invasion, and lymph node metastasis (all P < 0.05). The 5-year OS rate was also markedly higher in patients with weak CDK14 staining (17.19%) than in those with strong staining (6.98%), indicating that high CDK14 expression predicted poorer clinical characteristics and higher mortality rates.

Table 4.

Relationships between CDK14 and the clinical characteristics of patients with non-small cell lung cancer.

Variables	Strong CDK14 staining (n = 129)	Weak CDK14 staining (n = 64)	P
Age, years	53.27 ± 11.93	52.46 ± 10.82	0.648
BMI, kg/m2	22.85 ± 3.03	22.54 ± 3.02	0.509
Sex, male:female (%)	78 (60.47):51 (39.53)	41 (64.06):23 (35.94)	0.600
Current smoker, n (%)	69 (53.49)	38 (59.38)	0.401
TNM stage, n (%)			<0.001
I–II	48 (37.21)	41 (64.06)
III–IV	81 (62.79)	23 (35.94)
Pathological type, n (%)			0.018
Squamous cell carcinoma	78 (60.47)	28 (43.75)
Adenocarcinoma	51 (39.53)	36 (56.25)
Pleural invasion, n (%)	79 (76.70)	24 (37.50)	<0.001
Lymph node metastasis, n (%)	97 (75.19)	31 (48.44)	<0.001
5-year survival, n (%)	9 (6.98)	11 (17.19)	0.027

CDK14, cyclin-dependent kinase 14; BMI, body mass index.

Elevated CDK14 expression predicted shorter 5-year OS in patients with NSCLC

Finally, the association between CDK14 expression and 5-year OS was analyzed using Kaplan–Meier analysis. The median survival time was significantly shorter in patients with strong CDK14 staining than in those with weak CDK14 staining (P < 0.05). The Kaplan–Meier curve demonstrated that 5-year OS was markedly longer in patients with weak CDK14 staining than in those with strong CDK14 staining (P < 0.05 [log rank test], Figure 2). Logistic regression analysis identified TNM stage, pleural invasion, lymph node metastasis, CDK14 expression, and Ki67 expression as risk factors for 5-year OS in patients with NSCLC (Table 5).

Figure 2.

Kaplan–Meier curve of 5-year OS in patients with strong or weak CDK14 staining expression.

OS, overall survival; CDK14, cyclin-dependent kinase.

Table 5.

Logistic regression analysis of risk factors for 5-year overall survival in patients with non-small cell lung cancer.

	Wald	Odds ratio	95% confidence interval	P
Age	2.191	1.033	0.989–1.078	0.139
BMI	0.964	0.326	0.793–1.079	0.326
TNM stage	17.417	4.049	2.099–7.810	<0.001
Pathological type	3.412	2.703	0.941–7.765	0.064
Pleural invasion	1.746	0.174	0.049–0.616	0.006
Lymph node metastasis	5.440	0.088	0.011–0.679	0.019
CDK14	4.467	1.306	1.019–1.673	0.034
TTF-1	3.648	1.290	0.993–1.676	0.056
CK5/6	0.335	1.068	0.854–1.334	0.562
Ki67	4.341	1.315	1.016–1.703	0.037

BMI, body mass index; CDK14, cytokeratin-dependent kinase 14; TTF-1, thyroid transcription factor 1; CK5/6, cytokeratin 5/6.

Discussion

Currently, lung cancer remains a leading cause of cancer-related death, and its 5-year survival rate remains extremely low, especially among patients with advanced cancer. Most members of the CDK family are regarded as cancer promoters, including CDK14. However, the clinical significance of CDK14 in lung cancer has not been illuminated. In the present study, we demonstrated that CDK14 was overexpressed in NSCLC tumor tissues, and CDK14 overexpression portended poor prognosis.

First, this study demonstrated that CDK14 expression was elevated in tumor tissue in patients with NSCLC. CDKs inhibitors have already been used as anti-cancer agents in the clinic. In breast cancer, both CDK4 and CDK6 inhibitors were used as therapies, including the CDK 4/6 inhibitor palbociclib (PD0332991), which has been examined in phase II clinical trials. In gastric cancer, CDK2 promoted aerobic glycolysis and cell proliferation in gastric cancer cells by inhibiting SIRT5. Meanwhile, recent studies revealed that CDK15 was upregulated in lung cancer and that it suppressed cell apoptosis in breast cancer.[16,17] In our research, we also found that CDK14 was overexpressed in NSCLC and associated with poor prognosis.

The finding that CDK14 may serve as a cancer promoter in NSCLC is consistent with prior findings. In previous research, CDK14 activation led to the promotion of cell viability and metastasis in NSCLC cell lines. In addition, several studies have reported the role of CDK14 in cancer development. Ji et al. found that inhibition of CDK14 by miR-216 suppressed the proliferation, migration, and invasion of osteosarcoma cells. Another study demonstrated that suppression of CDK14 resulted in decreased breast cancer cell proliferation. All of these studies highlighted that CDK14 is a promoter of various cancers, including lung cancer. In this paper, we also demonstrated the elevated CDK14 expression in NSCLC predicted advanced disease stage, lymphatic metastasis, and shorter 5-year survival. Additionally, CDK14 expression was positively correlated with the expression of the tumor biomarkers TTF-1, CK5/6, and Ki67. In addition to the aforementioned studies, another study found that cigarette smoke downregulated CDK14 expression in testicular and lung cells, suggesting that CDK14 has different roles in normal lung tissue and lung cancer tissue. Thus, further studies are needed to reveal the underlying molecular mechanisms of CDK14 in NSCLC.

The present study had some limitations. First, the sample size was limited and all patients were recruited from a single center. Second, the underlying molecular mechanism of the effects of CDK14 in lung cancer is unclear. Thus, additional research is required for confirmation.

Conclusion

This observational study illustrated that CDK14 was overexpressed in patients with NSCLC, and CDK14 overexpression predicted poor prognoses and 5-year OS. The present study might provide new insights into the role of CDK14 in NSCLC development.