Expression and prognostic significance of cyclin‑dependent kinase inhibitor 1A in patients with resected gastric adenocarcinoma

  • Authors:
    • Youdong Lin
    • Xiaoting Wang
    • Yinghao Yu
    • Wei Liu
    • Feilai Xie
    • Xuenong Ouyang
    • Qiaojia Huang
  • View Affiliations

  • Published online on: September 29, 2017     https://doi.org/10.3892/ol.2017.7107
  • Pages:7473-7482
Metrics: HTML 0 views | PDF 0 views     Cited By (CrossRef): 0 citations

Abstract

Cyclin‑dependent kinase inhibitor 1A (CDKN1A) is an important cell cycleregulator, and has been identified to exhibit aberrant expression in various types of cancer tissues. However, the association between CDKN1A expression level and prognosis in patients with resected gastric adenocarcinoma (RGA) requires additional elucidation. In the present study, the CDKN1A expression profile in RGA tissues obtained from 217 patients were analyzed using immunohistochemistry. Its prognostic significance was evaluated by using the χ2 test, Kaplan‑Meier curves and the log‑rank test, and a multivariate Cox model analysis, during a median follow‑up time of 51 months. The results demonstrated that CDKN1A expression was significantly correlated with lymph node metastasis (LNM; P=0.001), recurrence (P<0.001) and overall survival (OS; P<0.001). In addition, the recurrence‑free survival (RFS) and OS times were significantly shorter in patients with low CDKN1A expression compared with those with high CDKN1A expression (RFS, 20 months vs. 69 months, P<0.001; and OS, 32 months vs. 70 months, P<0.001, respectively). Multivariate analysis additionally confirmed that low CDKN1A expression was significantly correlated with an increased risk of LNM (P=0.001), recurrence (P<0.001) and mortality (P<0.001). Therefore, these data suggest that low expression of CDKN1A has independent prognostic significance indicative of tumor progression and poor survival in patients with RGA. Evaluation of CDKN1A expression may assist in determining prognosis in patients with RGA.

Introduction

Gastric cancer is the most common digestive tract malignancy that affects ~1 million people per year worldwide (1), and >90% of these cases are gastric adenocarcinoma (GA). At present, surgery remains the primary method for treating patients with GA. Lymph node metastasis (LNM) usually indicates a poor prognosis of resected GA (RGA), as it is closely associated with widespread metastasis that will affect the survival of the patients (2,3). A reduced risk of metastasis and recurrence and improved survival time are the primary goals of cancer therapy. Therefore, it would be useful to identify a biomarker associated with LNM that may assist in predicting prognosis.

Cyclin-dependent kinase inhibitor 1A (CDKN1A; also known as p21), a cell cycle regulator which is encoded by the CDKN1A gene at chromosome 6p21.2, is primarily involved in mediating cell growth, proliferation, differentiation, DNA repair and apoptosis (46). Data from a number of studies have demonstrated that it assists in regulating tumor development by inducing cell cycle arrest, leading to a reduction in cancer cell growth rate (7,8). Aberrant expression of CDKN1A has been frequently observed in different types of cancer tissues (911). In the present study, in order to understand the association between CDKN1A expression and prognosis in patients with RGA, the correlation between the expression of CDKN1A and clinicopathological factors, LNM, recurrence rate, and survival in patients with RGA was evaluated by various statistical analyses.

Materials and methods

Patients

GA tissue samples for immunohistochemistry (IHC) were retrospectively collected from 217 patients who were diagnosed pathologically with primary GA and then underwent surgical resection between January 2003 and November 2008. Information was obtained from the medical records of each patient and follow-up records subsequent to treatment were monitored until March 2013. Survival data were obtained via telephone contact and the Social Security Death Index. Recurrence was diagnosed according to the criteria described in our previous study (12). Fluoropyrimidine-based regimens (fluoropyrimidine or fluoropyrimidine plus platinum) were used as first-line chemotherapy to treat all of the patients following resection of GA. The standards described in the guidelines of the International Union Against Cancer (UICC; 5th Edition) and the National Comprehensive Cancer Network's (NCCN) Clinical Practice Guidelines in Oncology (v.1.2011) (13,14) were used to determine the tumor-node-metastasis (TNM) stage and histological grade for each of the tissue samples used in the present study.

For comparison of the results of western blotting and IHC, GA tissue samples from 52 cases were collected from patients with primary GA who underwent surgical resection between March 2014 and December 2015.

The study was approved by the Ethics Committee of Fuzhou General Hospital (Fuzhou, China) and all patients provided informed consent.

IHC

IHC was used to determine the levels of CDKN1A in paraffin-embedded RGA (217 cases) and normal gastric tissues (30 cases). The anti-CDKN1A antibody was obtained from ZSGB-BIO (cat. no. TA808128; dilution, 1:150; Beijing, China). IHC was performed as described previously (12,15). The expression intensity of CDKN1A was scored from 0 to 3 according to the standards established by Kawata et al (16). Samples with a staining intensity of 0 or 1 were considered as low expression, and with staining intensity of 2 or 3 as high expression.

Hematoxylin and eosin (H&E) staining

H&E staining of paraffin-embedded normal gastric tissues and cancer tissues in patients with RGA were performed by a pathologist using an established methodology. The steps of H&E staining are briefly described as follows: i) Sections of tissue samples were dewaxed with xylene and washed with various concentrations of ethanol (100, 100, 95 and 80%); ii) stained with 0.5% hematoxylin for 5 min, washed with water for 10 min and then distilled water for 5 sec; iii) soaked in hydrochloric acid ethanol for 30 sec and then water for 15 min; iv) 0.5% eosinstaining for 2 min; v) conventional dehydration, transparency and neutral resins mount; vi) the results were observed using light microscopy (magnification, ×200). All steps were performed at room temperature (25°C).

Western blotting

As described previously (17), total protein from cancer tissues was extracted using RIPA lysis buffer (cat. no. P0013B) and quantified using the BCA Protein Assay kit (cat. no. P0012S), and western blotting was performed using a SDS-PAGE Gel preparation reagent kit [cat. no. P0012A: P0012A-1, 30% Acr-Bis (29:1); P0012A-2, 1M Tris-HCl, pH 8.8; P0012A-3, 10% SDS; P0012A-4, ammonium persulfate; P0012A-5, TEMED; P0012A-6, 1M Tris-HCl, pH 6.8; SDS-PAGE Sample Loading Buffer (5X), cat. no. P0015], blocking buffer (cat. no. P0023B) and enhanced chemiluminescence reagents (ECL; cat. no. P0018) (all of the above reagents were from Beyotime Institute of Biotechnology, Haimen, China). Briefly, the proteins (30 µg per lane) were subjected to SDS-PAGE gel electrophoresis (concentrated gel, 5%; separation gel, 12%) and then transferred to polyvinylidene membranes. Following incubation with blocking buffer at room temperature for 30 min, the membranes were incubated with primary polyclonal rabbit anti-human CDKN1A antibody (cat. no. SC-397; dilution, 1:300; ZSGB-BIO) or primary polyclonal rabbit anti-human GAPDH antibody (cat. no. 2118S; dilution, 1:1,000; Cell Signaling Technology, Inc., Danvers, MA, USA) overnight at 4°C. Following three washes with TBST the membranes were incubated with secondary goat anti-rabbit antibody IgG-horseradish peroxidase (cat. no. M21002M; dilution, 1:2,000; abMart Company, Shanghai, China) for 4 h at room temperature. Finally, the bands were visualized using ECL and analyzed with a chemiluminescence imager system (Image Quant LAS4000 mini system; GE Healthcare, Chicago, IL, USA).

Statistical analysis

SPSS 17.0 statistical software (SPSS, Inc., Chicago, IL, USA) was used to conduct data analysis, including the correlation between CDKN1A expression and the clinicopathological characteristics and prognosis of patients. The methods used in this analysis were the χ2 test, and Kaplan-Meier and log-rank analyses of recurrence-free survival (RFS) and overall survival (OS) times in months. Additionally, univariate and multivariate Cox models were used to assess the prognostic ability of known unfavorable pathological factors, and the correlation between CDKN1A expression profiles and recurrence or survival. The agreement between CDKN1A expression detected by western blot and immunohistochemical analyses was evaluated by Spearman's rank correlation analysis. All tests were two-sided, and P<0.05 was considered to indicate a statistically significant difference.

Results

Clinicopathological characteristics of patients

The clinicopathological characteristics of the 217 RGA patients are summarized in Table I. All patients had received fluoropyrimidine or fluoropyrimidine plus platinum agents as first-line chemotherapy for an average of 4 cycles following resection. The patients comprised 153 males and 64 females; 135 patients were <60 years old, and 82 patients were ≥60 years old, the median age was 60 years. At the time of resection, 147 patients (67.7%) presented with LNM. In 145 cases (66.8%), the tumors were moderately or well-differentiated, while the remaining 72 cases (33.2%) exhibited poorly differentiated histology. By the end of follow-up, 126 had developed recurrent disease, and 116 patients had succumbed to the disease. The median times to recurrence and mortality were 41 and 54 months, respectively. The median time of observation was 51 months.

Table I.

Patient characteristics (n=217).

Table I.

Patient characteristics (n=217).

Cases

Characteristicsn%
Age, years
  <6013562.2
  ≥60  8237.8
Sex
  Male15370.5
  Female  6429.5
Histological grade
  Well/moderately differentiated14566.8
  Poorly differentiated  7233.2
Gross findings
  Apophysis  177.8
  Invasion20092.2
Tumor site
  Proximal  4520.7
  Mid/distal17279.3
Lymph node metastasis
  No metastasis  7032.3
  Metastasis14767.7
T stage
  T1/T210146.5
  T3/T411653.5
Postoperative chemotherapy
  Fluoropyrimidines  6228.6
  Fluoropyrimidines + platinum15571.4
CDKN1A expression
  Low  9945.6
  High11854.4
Survival
  Alive   10146.5
  Succumbed11653.5
Recurrence
  No  9141.9
  Yes12658.1
CDKN1A expression and correlation between CDKN1A expression and clinicopathological characteristics

CDKN1A exhibited differential expression patterns in RGA and normal gastric tissues. In normal gastric tissues, CDKN1A protein was located mainly in the cell nuclei, whereas CDKN1A protein exhibited nuclear and cytoplasmic expression in RGA tissues. Representative IHC results of the different intensities of CDKN1A expression in RGA tissues and in normal gastric tissues are presented in Fig. 1, along with their corresponding H&E staining results. Representative western blot results are shown in Fig. 2. Resected tumor tissues from 118 cases (54.4%) of RGA exhibited high CDKN1A expression, whereas those from 99 patients (45.6%) demonstrated low CDKN1A expression (Table I). As summarized in Table II, CDKN1A expression was significantly associated with LNM (P=0.001), recurrence (P<0.001), and survival (P<0.001), as determined by χ2 test. No significant associations between CDKN1A expression and other clinicopathological characteristics, including age, gender, histological grade, gross pathology, tumor site, and postoperative chemotherapy, were observed (Table II).

Table II.

Association between CDKN1A expression and characteristics of patients with resected gastric adenocarcinoma (n=217).

Table II.

Association between CDKN1A expression and characteristics of patients with resected gastric adenocarcinoma (n=217).

CDKN1A expression, n (%)

CharacteristicsLowHighP-value
Age, years 0.072
  <6068 (50.4)67 (49.6)
  ≥6031 (37.8)51 (62.2)
Sex 0.720
  Male71 (46.4)82 (53.6)
  Female28 (43.8)36 (56.2)
Histological grade 0.593
  Well/moderately differentiated68 (46.9)77 (53.1)
  Poorly differentiated31 (43.1)41 (56.9)
Gross findings 0.701
  Apophysis7 (41.2)10 (58.8)
  Invasion92 (46.0)108 (54.0)
Tumor site 0.621
  Proximal22 (48.9)23 (51.1)
  Mid/distal77 (44.8)95 (55.2)
Lymph node metastasis 0.001a
  No metastasis20 (28.6)50 (71.4)
  Metastasis79 (53.7)68 (46.3)
T stage 0.053
  T1/T239 (38.6)62 (61.4)
  T3/T460 (51.7)56 (48.3)
Postoperative chemotherapy 0.085
  Fluoropyrimidines34 (54.8)28 (45.2)
  Fluoropyrimidines + platinum65 (41.9)90 (58.1)
Survival <0.001a
  Alive25 (24.8)76 (75.2)
  Succumbed74 (63.8)42 (36.2)
Recurrence <0.001a
  No20 (22.0)71 (78.0)
  Yes79 (62.7)47 (37.3)

a Statistically significant (P<0.05). CDKN1A, cyclin-dependent kinase inhibitor 1A.

Identification of prognostic factors associated with OS

The 5-year OS rate of the entire patient group was 46.5% (Table I). Data from the χ2 and log-rank tests indicated that patients with low CDKN1A expression exhibited a significantly poorer survival rate (25.3%; χ2 P<0.001) and lower median OS time (log-rank P<0.001; Tables II and III) compared with patients with high CDKN1A expression. As demonstrated in Table III, other factors identified by χ2 and log-rank tests to be associated with poor survival included LNM (P<0.001 and P<0.001, respectively), advanced T stage (P<0.001 and P<0.001, respectively), and postoperative chemotherapy with fluoropyrimidine only (P=0.008 and P=0.002, respectively). The results of the Kaplan-Meier analysis demonstrated that patients with low CDKN1A expression exhibited a significantly shorter OS time compared with those with high CDKN1A expression (median, 32 months vs. 70 months, P<0.001; Fig. 3A).

Table III.

Association between OS and characteristics of patients with resected gastric adenocarcinoma (n=217).

Table III.

Association between OS and characteristics of patients with resected gastric adenocarcinoma (n=217).

Survival

CharacteristicsAlive (%)Deceased (%) P-valueaMedian OS time (months) P-valueb
Age, years 0.743 0.783
  <6064 (47.4)71 (52.6) 57.0
  ≥6037 (45.1)45 (54.9) 48.0
Sex 0.718 0.611
  Male70 (45.8)83 (54.2) 51.0
  Female31 (48.4)33 (51.6) 53.0
Histological grade 0.468 0.417
  Well/moderately differentiated70 (48.3)75 (51.7) 54.0
  Poorly differentiated31 (43.1)41 (56.9) 41.0
Gross findings 0.290 0.229
  Apophysis10 (58.8)7 (41.2) 63.0
  Invasion91 (45.5)109 (54.5) 47.0
Tumor site 0.751 0.733
  Proximal20 (44.4)25 (55.6) 51.0
  Mid/distal81 (47.1)91 (52.9) 53.0
Lymph node metastasis <0.001c <0.001c
  No metastasis46 (65.7)24 (34.3) 76.0
  Metastasis55 (37.4)92 (62.6) 39.0
T stage <0.001c <0.001c
  T1/T263 (62.4)38 (37.6) 69.0
  T3/T438 (32.8)78 (67.2) 35.0
Postoperative chemotherapy 0.008c 0.002c
  Fluoropyrimidines20 (32.3)42 (67.7) 33.0
  Fluoropyrimidines + platinum81 (52.3)74 (47.7) 63.0
CDKN1A expression <0.001c <0.001c
  Low25 (25.3)74 (74.7) 32.0
  High76 (64.4)42 (35.6) 70.0

a P-value was obtained from the χ2 test

b P-value was obtained from the log-rank test

c statistically significant (P<0.05). OS, overall survival; CDKN1A, cyclin-dependent kinase inhibitor 1A.

Results from the univariate and multivariate Cox analyses additionally confirmed that positive LNM (P<0.001 and P=0.006, respectively), advanced T stage (T3/T4) (P<0.001 and P=0.001, respectively), postoperative chemotherapy with fluoropyrimidine only (P=0.003 and P=0.003, respectively) and low CDKN1A expression (P<0.001 and P<0.001, respectively) were significantly associated with increased risk of mortality (Table IV). Taken together, these data suggest that CDKN1A was a statistically significant independent prognostic significance for OS in patients with RGA.

Table IV.

Univariate and multivariate Cox regression analyses of overall survival.

Table IV.

Univariate and multivariate Cox regression analyses of overall survival.

Univariate analysisMultivariate analysis


CharacteristicsHR (95% CI)P-valueHR (95% CI)P-value
Age, years 0.785 0.366
  <601.00 1.00
  ≥601.05 (0.73–1.53) 1.20 (0.81–1.78)
Sex 0.614 0.188
  Male1.00 1.00
  Female0.90 (0.60–1.35) 0.75 (0.49–1.15)
Histological grade 0.422 0.252
  Well/moderately differentiated1.00 1.00
  Poorly differentiated1.17 (0.80–1.71) 1.26 (0.85–1.89)
Gross findings 0.237 0.441
  Apophysis1.00 1.00
  Invasion1.59 (0.74–3.41) 1.36 (0.62–2.98)
Tumor site 0.735 0.462
  Proximal1.00 1.00
  Mid/distal0.93 (0.60–1.44) 0.84 (0.53–1.34)
Lymph node metastasis <0.001a 0.006a
  No metastasis1.00 1.00
  Metastasis2.40 (1.53–3.76) 1.94 (1.21–3.14)
T stage <0.001a 0.001a
  T1/T21.00 1.00
  T3/T42.38 (1.61–3.51) 2.03 (1.36–3.03)
Postoperative chemotherapy 0.003a 0.003a
  Fluoropyrimidines1.00 1.00
  Fluoropyrimidines + platinum0.56 (0.38–0.81) 0.55 (0.37–0.82)
CDKN1A expression <0.001a <0.001a
  Low1.00 1.00
  High0.33 (0.22–0.48) 0.39 (0.26–0.58)

a Statistically significant (P<0.05). HR, hazard ratio; CI, confidence interval; CDKN1A, cyclin-dependent kinase inhibitor 1A.

Identification of prognostic factors associated with recurrence risk

The overall recurrence rate of the whole patient cohort was 58.1% (Table I). Data from the χ2 and log-rank tests indicated that patients with low CDKN1A expression exhibited a significantly higher recurrence rate (79.8%; χ2 P<0.001) and lower median recurrence-free survival (log-rank P<0.001) compared with those with high CDKN1A expression (Tables II and V). As demonstrated in Table V, other factors identified by χ2 and log-rank tests to be associated with a higher recurrence rate included LNM (P<0.001 and P<0.001, respectively), and T3/T4 stage (P<0.001 and P<0.001, respectively). The results of the Kaplan-Meier analysis indicated that patients with low CDKN1A expression exhibited a significantly shorter RFS time compared with those with high CDKN1A expression (median, 20 vs. 69 months, P<0.001; Fig. 3B).

Table V.

Association between recurrence and characteristics of patients with resected gastric carcinoma (n=217).

Table V.

Association between recurrence and characteristics of patients with resected gastric carcinoma (n=217).

Recurrence, n (%)

CharacteristicsNoYes P-valueaMedian RFS time (months) P-valueb
Age, years 0.647 0.461
  <6055 (40.7)80 (59.3) 35.0
  ≥6036 (43.9)46 (56.1) 42.0
Sex 0.961 0.876
  Male64 (41.8)89 (58.2) 39.0
  Female27 (42.2)37 (57.8) 41.0
Histological grade 0.727 0.684
  Well/moderately differentiated62 (42.8)83 (57.2) 41.0
  Poorly differentiated29 (40.3)43 (59.7) 36.0
Gross findings 0.338 0.249
  Apophysis9 (52.9)8 (47.1) 63.0
  Invasion82 (41.0)118 (59.0) 36.0
Tumor site 0.330 0.428
  Proximal16 (35.6)29 (64.4) 24.0
  Mid/distal75 (43.6)97 (56.4) 40.0
Lymph node metastasis <0.001c <0.001c
  No metastasis43 (61.4)27 (38.6) 70.5
  Metastasis48 (32.7)99 (67.3) 24.0
T stage <0.001c <0.001c
  T1/T256 (55.4)45 (44.6) 68.0
  T3/T435 (30.2)81 (69.8) 24.0
Postoperative chemotherapy 0.068 0.035c
  Fluoropyrimidines20 (32.3)42 (67.7) 24.0
  Fluoropyrimidines + platinum71 (45.8)84 (54.2) 53.0
CDKN1A expression <0.001c <0.001c
  Low20 (20.2)79 (79.8) 20.0
  High71 (60.2)47 (39.8) 69.0

a P-value was obtained from the χ2 test

b P-value was obtained from the log-rank test

c statistically significant (P<0.05). RFS, recurrence-free survival; CDKN1A, cyclin-dependent kinase inhibitor 1A.

Results from the univariate and multivariate Cox analyses demonstrated that LNM (P<0.001 and P=0.001, respectively), advanced T stage (P<0.001 and P=0.003, respectively), postoperative chemotherapy with fluoropyrimidine only (P=0.038 and P=0.007) and low CDKN1A expression (P<0.001 and P<0.001) were significantly associated with increased risk of recurrence (Table VI). Taken together, these data suggest that CDKN1A was a statistically significant independent prognostic significance for recurrence risk in patients with RGA.

Table VI.

Univariate and multivariate Cox regression analysis of recurrence.

Table VI.

Univariate and multivariate Cox regression analysis of recurrence.

Univariate analysisMultivariate analysis


CharacteristicsHR (95% CI)P-valueHR (95% CI)P-value
Age, years 0.466 0.733
  <601.00 1.00
  ≥600.87 (0.61–1.26) 0.94 (0.64–1.37)
Sex 0.877 0.165
  Male1.00 1.00
  Female0.97 (0.66–1.42) 0.75 (0.50–1.13)
Histological grade 0.687 0.507
  Well/moderately differentiated1.00 1.00
  Poorly differentiated1.08 (0.75–1.56) 1.14 (0.77–1.68)
Gross findings 0.257 0.470
  Apophysis1.00 1.00
  Invasion1.51 (0.74–3.10) 1.31 (0.63–2.75)
Tumor site 0.434 0.315
  Proximal1.00 1.00
  Mid/distal0.85 (0.56–1.28) 0.80 (0.51–1.24)
Lymph node metastasis <0.001a 0.001a
  No metastasis1.00 1.00
  Metastasis2.54 (1.66–3.90) 2.09 (1.33–3.27)
T stage <0.001a 0.003a
  T1/T21.00 1.00
  T3/T42.16 (1.50–3.13) 1.77 (1.21–2.58)
Postoperative chemotherapy 0.038a 0.007a
  Fluoropyrimidines1.00 1.00
  Fluoropyrimidines + platinum0.67 (0.46–0.98) 0.58 (0.39–0.87)
CDKN1A expression <0.001a <0.001a
  Low1.00 1.00
  High0.31 (0.22–0.45) 0.38 (0.26–0.56)

a Statistically significant (P<0.05). HR, hazard ratio; CI, confidence interval; CDKN1A, cyclin-dependent kinase inhibitor 1A.

Identification of risk factors associated with LNM

At the time of surgical resection, 79 of 99 patients (79.8%) with low CDKN1A expression presented with LNM, compared with 68 of 118 patients (57.6%) with high CDKN1A expression, a difference that was identified to be significant (P=0.001 by χ2 test; Tables II and VII). Univariate and multivariate logistic regression analyses additionally demonstrated that low CDKN1A expression was significantly associated with increased risk of LNM (P=0.001 and P=0.001, respectively). Other factors, including gross pathology (P=0.02 and P=0.035 from univariate and multivariate analyses, respectively) and advanced T stage (T3/T4; P=0.001 and P=0.005, from univariate and multivariate analyses, respectively) were also significantly associated with LNM (Table VII). Taken together, these data suggest that CDKN1A was a statistically significant independent risk factor for LNM in patients with RGA.

Table VII.

Univariate and multivariate analyses of variables correlated with LNM.

Table VII.

Univariate and multivariate analyses of variables correlated with LNM.

LNM, n (%)Univariate analysisMultivariate analysis



CharacteristicsNoYesOR (95% CI)P-valueOR (95% CI)P-value
Age, years 0.643 0.206
  <6042 (31.1)93 (68.9)1.00 1.00
  ≥6028 (34.1)54 (65.9)0.87 (0.49–1.56)0.99 (0.52–1.91)
Sex 0.247 0.991
  Male53 (34.6)100 (65.4)1.00 1.00
  Female17 (26.6)47 (73.4)1.47 (0.77–2.80)1.58 (0.78–3.19)
Histological grade 0.584 0.363
  Well/moderately differentiated45 (31.0)100 (69.0)1.00 1.00
  Poorly differentiated25 (34.7)47 (65.3)0.85 (0.47–1.54)0.73 (0.37–1.43)
Gross findings 0.020a 0.035a
  Apophysis10 (58.8)7 (41.2)1.00 1.00
  Invasion60 (30.0)140 (70.0)3.33 (1.21–9.17) 3.31 (1.09–10.02)
Tumor site 0.587 0.736
  Proximal13 (28.9)32 (71.1)1.00 1.00
  Mid/distal57 (33.1)115 (66.9)0.82 (0.40–1.68) 0.87 (0.40–1.92)
T stage 0.001a 0.005a
  T1/T244 (43.6)57 (56.4)1.00 1.00
  T3/T426 (22.4)90 (77.6)2.67 (1.49–4.81) 2.44 (1.31–4.53)
Postoperative chemotherapy 0.200 0.102
  Fluoropyrimidines24 (38.7)38 (61.3)1.00 1.00
  Fluoropyrimidines + platinum46 (29.7)109 (70.3)1.50 (0.81–2.77) 1.78 (0.89–3.55)
CDKN1A expression 0.001a 0.001a
  Low20 (20.2)79 (79.8)1.00 1.00
  High50 (42.4)68 (57.6)0.34 (0.19–0.64) 0.33 (0.17–0.64)

a Statistically significant (P<0.05). LNM, lymph node metastasis; OR, odds ratio; CDKN1A, cyclin-dependent kinase 1A.

Correlation between western blot and IHC analyses for detection of CDKN1A

As presented in the representative results of Fig. 2, CDKN1A protein levels in tissues from 52 patients with RGA detected by western blotting and IHC exhibited a good correlation (r=0.872, P<0.001), which was obtained by scanning the western-blot bands with the gel analysis system to obtain the gray value, and then performing Spearman's rank correlation analysis to analyze the correlation between the gray values and their corresponding IHC scores so as to plot the scatter plot (Fig. 4). These data suggest that CDKN1A expression in RGA tissues may be determined by either western blotting or immunohistochemical analysis.

Discussion

The results of the present study demonstrated that low CDKN1A expression in GA tissues was associated with LNM and poor prognoses, as indicated by the shorter survival times and increased recurrence risks in these patients. Thus, CDKN1A appears to have independent prognostic significance in patients with RGA.

Previously published data have demonstrated that CDKN1A is a tumor suppressor due to its ability to induce cell cycle arrest (1821). Confirming this role, downregulation of CDKN1A has been identified in different cancer tissues, and has been associated with poor prognoses in various types of human cancer, such as pancreatic and colon cancer, primary hepatocellular carcinoma, early-stage human papillomavirus-associated lung cancer and ovarian cancer (4,2227). Khalili et al (28) revealed that low CDKN1A expression level may be associated with gastric cancer initiation and progression. However, there have been conflicting results in terms of the CDKN1A expression profiles in different cancer tissues. Certain studies have identified that CDKN1A overexpression in cancer tissues is associated with tumor aggressiveness and poor survival: Dai et al (29) suggested that CDKN1A expression in breast cancer tissues was correlated with LNM and poor survival of patients with breast cancer, and increased MDA-MB231 breast cancer cell migration and invasion via transforming growth factor β-mediated mothers against decapentaplegic homolog 3 acetylation. In addition to the aforementioned studies (4,2227), there have been other investigations indicating that low expression of CDKN1A in various cancer tissues is associated with poor prognosis (30,31). The molecular mechanism by which different CDKN1A expression levels affect prognoses in patients with different types of cancer remains unclear and requires additional investigation. The present study identified that low expression of CDKN1A in RGA tissues was associated with poor prognosis in patients with gastric cancer.

As CDKN1A is a tumor suppressor, we hypothesize that epigenetic changes, such as promoter methylation, may be one of the mechanisms underlying the low CDKN1A expression in RGA tissues; promoter methylation is capable of inhibiting gene transcript and protein expression, as identified by Watanabe et al (32), who suggested that abnormal genomic methylation may be involved in the reduced expression of CDKN1A in adult T-cell leukemia/lymphoma (32). A study conducted by Nie et al (33) indicated that long noncoding RNAs (lncRNA) may also serve a role in regulation of CDKN1A transcription; the authors demonstrated that the lncRNA antisense noncoding RNA in the INK4 locus was upregulated in non-small cell lung cancer (NSCLC) tissues, and that this lncRNA promoted NSCLC cell proliferation and inhibited NSCLC cell apoptosis through the inhibition of Krüppel-like factor 2 and CDKN1A transcription (33). These studies additionally demonstrate that CDKN1A is a key regulator of the cell cycle, and may be a target for cancer treatment.

In conclusion, the present study indicated, via multiple statistical analyses, that low expression level of CDKN1A in RGA tissues was significantly associated with LNM, a shorter survival time and a high recurrence rate and risk of mortality. CDKN1A demonstrated important prognostic significance, and may be used as an independent prognostic factor for patients with RGA.

Acknowledgements

The present study was supported by grants from the National NaturalScience Foundation of China (grant no. 81372788), the Medical Scientific Research Key Foundation of Nanjing Command (grant no. 11Z032), and the Natural Science Foundation of Fujian Province Grant (grant no. 2014J01427), all awarded to Professor Qiaojia Huang; the Medical Scientific Innovation Foundation of Nanjing Command (grant no. 2013-MS122), awarded to Professor Yinghao Yu; and the National Natural Science Foundation of China (grant no. 81274002), awarded to Professor Xuenong Ouyang.

References

1 

Jemal A, Bray F, Center MM, Ferlay J, Ward E and Forman D: Global cancer statistics. CA Cancer J Clin. 61:69–90. 2011. View Article : Google Scholar : PubMed/NCBI

2 

Cunningham D, Allum WH, Stenning SP, Thompson JN, Van de Velde CJ, Nicolson M, Scarffe JH, Lofts FJ, Falk SJ, Iveson TJ, et al: Perioperative chemotherapy versus surgery alone for resectable gastroesophageal cancer. N Engl J Med. 355:11–20. 2006. View Article : Google Scholar : PubMed/NCBI

3 

Liu C, Lu P, Lu Y, Xu H, Wang S and Chen J: Clinical implications of metastatic lymph node ratio in gastric cancer. BMC Cancer. 7:2002007. View Article : Google Scholar : PubMed/NCBI

4 

Ohta K, Hoshino H, Wang J, Ono S, Iida Y, Hata K, Huang SK, Colquhoun S and Hoon DS: MicroRNA-93 activates c-Met/PI3K/Akt pathway activity in hepatocellular carcinoma by directly inhibiting PTEN and CDKN1A. Oncotarget. 6:3211–3324. 2015. View Article : Google Scholar : PubMed/NCBI

5 

Cazzalini O, Scovassi AI, Savio M, Stivala LA and Prosperi E: Multiple roles of the cell cycle inhibitor p21 (CDKN1A) in the DNA damage response. Mutat Res. 704:12–20. 2010. View Article : Google Scholar : PubMed/NCBI

6 

Prives C and Gottifredi V: The p21 and PCNA partnership: A new twist for an old plot. Cell Cycle. 7:3840–3846. 2008. View Article : Google Scholar : PubMed/NCBI

7 

Andries V, Vandepoele K, Staes K, Berx G, Bogaert P, Van Isterdael G, Ginneberge D, Parthoens E, Vandenbussche J, Gevaert K, et al: NBPF1, a tumor suppressor candidate in neuroblastoma, exerts growth inhibitory effects by inducing a G1 cell cycle arrest. BMC Cancer. 15:3912015. View Article : Google Scholar : PubMed/NCBI

8 

Park C, Jeong NY, Kim GY, Han MH, Chung IM, Kim WJ, Yoo YH and Choi YH: Momilactone B induces apoptosis and G1 arrest of the cell cycle in human monocytic leukemia U937 cells through downregulation of pRB phosphorylation and induction of the cyclin-dependent kinase inhibitor p21Waf1/Cip1. Oncol Rep. 31:1653–1660. 2014. View Article : Google Scholar : PubMed/NCBI

9 

Wei CY, Tan QX, Zhu X, Qin QH, Zhu FB, Mo QG and Yang WP: Expression of CDKN1A/p21 and TGFBR2 in breast cancer and their prognostic significance. Int J Clin Exp Pathol. 8:14619–14629. 2015.PubMed/NCBI

10 

Li S, Wang C, Yu X, Wu H, Hu J, Wang S and Ye Z: miR-3619-5p inhibits prostate cancer cell growth by activating CDKN1A expression. Oncol Rep. 37:241–248. 2017. View Article : Google Scholar : PubMed/NCBI

11 

Tang H, Wu Y, Liu M, Qin Y, Wang H, Wang L, Li S, Zhu H, He Z, Luo J, et al: SEMA3B improves the survival of patients with esophageal squamous cell carcinoma by upregulating p53 and p21. Oncol Rep. 36:900–908. 2016. View Article : Google Scholar : PubMed/NCBI

12 

Wang X, Lin Y, Lan F, Yu Y, Ouyang X, Wang X, Huang Q, Wang L, Tan J and Zheng F: A GG allele of 3′-side AKT1 SNP is associated with decreased AKT1 activation and better prognosis of gastric cancer. J Cancer Res Clin Oncol. 140:1399–1411. 2014. View Article : Google Scholar : PubMed/NCBI

13 

Yoo CH, Noh SH, Kim YI and Min JS: Comparison of prognostic significance of nodal staging between old (4th edition) and new (5th edition) UICC TNM classification for gastric carcinoma. International union against cancer. World J Surg. 23:492–497. 1999. View Article : Google Scholar : PubMed/NCBI

14 

Ward JH: NCCN Guidelines and the International community. J Natl Compr Canc Netw. 9:133–134. 2011. View Article : Google Scholar : PubMed/NCBI

15 

Wang X, Lin Y, Lan F, Yu Y, Ouyang X, Liu W, Xie F, Wang X and Huang Q: BAX and CDKN1A polymorphisms correlated with clinical outcomes of gastric cancer patients treated with postoperative chemotherapy. Med Oncol. 31:2492014. View Article : Google Scholar : PubMed/NCBI

16 

Kawata N, Tsuchiya N, Horikawa Y, Inoue T, Tsuruta H, Maita S, Satoh S, Mitobe Y, Narita S and Habuchi T: Two surviving polymorphisms are cooperatively associated with bladder cancer susceptibility. Int J Cancer. 129:1872–1880. 2011. View Article : Google Scholar : PubMed/NCBI

17 

Huang Q, Huang Q, Chen W, Wang L, Lin W, Lin J and Lin X: Identification of transgelin as a potential novel biomarker for gastric adenocarcinoma based on proteomics technology. J Cancer Res Clin Oncol. 134:1219–1227. 2008. View Article : Google Scholar : PubMed/NCBI

18 

Galeano F, Rossetti C, Tomaselli S, Cifaldi L, Lezzerini M, Pezzullo M, Boldrini R, Massimi L, Di Rocco CM, Locatelli F, et al: ADAR2-editing activity inhibits glioblastoma growth through the modulation of the CDC14B/Skp2/p21/p27 axis. Oncogene. 32:998–1009. 2013. View Article : Google Scholar : PubMed/NCBI

19 

Morton JP, Jamieson NB, Karim SA, Athineos D, Ridgway RA, Nixon C, McKay CJ, Carter R, Brunton VG, Frame MC, et al: LKB1 haploinsufficiency cooperates with Kras to promote pancreatic cancer through suppression of p21-dependent growth arrest. Gastroenterology. 139:586–597. 2010. View Article : Google Scholar : PubMed/NCBI

20 

Brown SG, Knowell AE, Hunt A, Patel D, Bhosle S and Chaudhary J: Interferon inducible antiviral MxA is inversely associated with prostate cancer and regulates cell cycle, invasion and Docetaxel induced apoptosis. Prostate. 75:266–279. 2015. View Article : Google Scholar : PubMed/NCBI

21 

Koh DI, Han D, Ryu H, Choi WI, Jeon BN, Kim MK, Kim Y, Kim JY, Parry L, Clarke AR, et al: KAISO, a critical regulator of p53-mediated transcription of CDKN1A and apoptotic genes. Proc Natl Acad Sci USA. 111:pp. 15078–15083. 2014, View Article : Google Scholar : PubMed/NCBI

22 

Sun Y, Yang S, Sun N and Chen J: Differential expression of STAT1 and p21 proteins predicts pancreatic cancer progression and prognosis. Pancreas. 43:619–623. 2014. View Article : Google Scholar : PubMed/NCBI

23 

Belt EJ, Brosens RP, Delis-van Diemen PM, Bril H, Tijssen M, van Essen DF, Heymans MW, Beliën JA, Stockmann HB, Meijer S, et al: Cell cycle proteins predict recurrence in stage II and III colon cancer. Ann Surg Oncol. 19 Suppl 3:S682–S692. 2012. View Article : Google Scholar : PubMed/NCBI

24 

Mitomi H, Ohkura Y, Fukui N, Kanazawa H, Kishimoto I, Nakamura T, Yokoyama K, Sada M, Kobayashi K, Tanabe S, et al: P21WAF1/CIP1 expression in colorectal carcinomas is related to Kras mutations and prognosis. Eur J Gastroenterol Hepatol. 19:883–889. 2007. View Article : Google Scholar : PubMed/NCBI

25 

Wu DW, Liu WS, Wang J, Chen CY, Cheng YW and Lee H: Reduced p21 (WAF1/CIP1) via alteration of p53-DDX3 pathway is associated with poor relapse-free survival in early-stage human papillomavirus-associated lung cancer. Clin Cancer Res. 17:1895–1905. 2011. View Article : Google Scholar : PubMed/NCBI

26 

Young TW, Rosen DG, Mei FC, Li N, Liu J, Wang XF and Cheng X: Up-regulation of tumor susceptibility gene 101 conveys poor prognosis through suppression of p21 expression in ovarian cancer. Clin Cancer Res. 13:3848–3854. 2007. View Article : Google Scholar : PubMed/NCBI

27 

Buchynska LG, Nesina IP, Yurchenko NP, Bilyk OO, Grinkevych VN and Svintitsky VS: Expression of p53, p21WAF1/CIP1, p16INK4A and Ki-67 proteins in serous ovarian tumors. Exp Oncol. 29:49–53. 2007.PubMed/NCBI

28 

Khalili M, Vasei M, Khalili D, Alimoghaddam K, Sadeghizadeh M and Mowla SJ: Downregulation of the genes involved in reprogramming (SOX2, c-MYC, miR-302, miR-145, and P21) in gastric adenocarcinoma. J Gastrointest Cancer. 46:251–258. 2015. View Article : Google Scholar : PubMed/NCBI

29 

Dai M, Al-Odaini AA, Arakelian A, Rabbani SA, Ali S and Lebrun JJ: A novel function for p21Cip1 and acetyltransferase p/CAF as critical transcriptional regulators of TGF β-mediated breast cancer cell migration and invasion. Breast Cancer Res. 14:R1272012. View Article : Google Scholar : PubMed/NCBI

30 

Zhang H, Zhang X, Ji S, Hao C, Mu Y, Sun J and Hao J: Sohlh2 inhibits ovarian cancer cell proliferation by upregulation of p21 and downregulation of cyclin D1. Carcinogenesis. 35:1863–1871. 2014. View Article : Google Scholar : PubMed/NCBI

31 

Liu J, Hu Y, Hu W, Xie X, Bella Ela A, Fu J and Rao D: Expression and prognostic relevance of p21WAF1 in stage III esophageal squamous cell carcinoma. Dis Esophagus. 25:67–71. 2012. View Article : Google Scholar : PubMed/NCBI

32 

Watanabe M, Nakahata S, Hamasaki M, Saito Y, Kawano Y, Hidaka T, Yamashita K, Umeki K, Taki T, Taniwaki M, et al: Downregulation of CDKN1A in adult T-cell leukemia/lymphoma despite overexpression of CDKN1A in human T-lymphotropic virus 1-infected cell lines. J Virol. 84:6966–6977. 2010. View Article : Google Scholar : PubMed/NCBI

33 

Nie FQ, Sun M, Yang JS, Xie M, Xu TP, Xia R, Liu YW, Liu XH, Zhang EB, Lu KH, et al: Long noncoding RNA ANRIL promotes non-small cell lung cancer cell proliferation and inhibits apoptosis by silencing KLF2 and P21 expression. Mol Cancer Ther. 14:268–277. 2015. View Article : Google Scholar : PubMed/NCBI

Related Articles

Journal Cover

December 2017
Volume 14 Issue 6

Print ISSN: 1792-1074
Online ISSN:1792-1082

Sign up for eToc alerts

Recommend to Library

Copy and paste a formatted citation
APA
Lin, Y., Wang, X., Yu, Y., Liu, W., Xie, F., Ouyang, X., & Huang, Q. (2017). Expression and prognostic significance of cyclin‑dependent kinase inhibitor 1A in patients with resected gastric adenocarcinoma. Oncology Letters, 14, 7473-7482. https://doi.org/10.3892/ol.2017.7107
MLA
Lin, Y., Wang, X., Yu, Y., Liu, W., Xie, F., Ouyang, X., Huang, Q."Expression and prognostic significance of cyclin‑dependent kinase inhibitor 1A in patients with resected gastric adenocarcinoma". Oncology Letters 14.6 (2017): 7473-7482.
Chicago
Lin, Y., Wang, X., Yu, Y., Liu, W., Xie, F., Ouyang, X., Huang, Q."Expression and prognostic significance of cyclin‑dependent kinase inhibitor 1A in patients with resected gastric adenocarcinoma". Oncology Letters 14, no. 6 (2017): 7473-7482. https://doi.org/10.3892/ol.2017.7107