p19^INK4d mRNA and protein expression as new prognostic factors in ovarian cancer patients

Abstract

p19^INK4d (CDKN2D) is a negative regulator of the cell cycle. Little is known of its role in cancer development and prognosis. We aimed to evaluate the clinical significance of p19^INK4d expression in ovarian carcinomas with respect to the TP53 accumulation status, as well as the frequency of CDKN2D mutations. p19^INK4d and TP53 expression was evaluated immunohistochemically in 445 ovarian carcinomas: 246 patients were treated with platinum–cyclophosphamide (PC/PAC), while 199 were treated with taxane–platinum agents (TP). CDKN2D gene expression (mRNA) was examined in 106 carcinomas, while CDKN2D mutations in 68 tumors. Uni- and multivariate statistical analyses (logistic regression and the Cox proportional hazards model) were performed for patient groups divided according to the chemotherapeutic regimen administered, and in subgroups with and without TP53 accumulation. High p19^INK4d expression increased the risk of death, but only in patients with the TP53-negative carcinomas (HR 1.61, P = 0.049 for PC/PAC-treated patients, HR 2.00, P = 0.015 for TP-treated patients). This result was confirmed by the mRNA analysis (HR 4.24, P = 0.001 for TP-treated group). High p19^INK4d protein expression associated with adverse clinicopathological factors. We found no alterations in the CDKN2D gene; the c.90C>G (p.R30R; rs1968445) polymorphism was detected in 10% of tumors. Our results suggest that p19^INK4d expression is a poor prognostic factor in ovarian cancer patients. Analyses of tumor groups according to the TP53 accumulation status facilitate the identification of cancer biomarkers.

Keywords: :

• CDKN2D
• chemotherapy
• immunohistochemistry
• ovarian cancer
• p19^INK4d
• prognostic factor
• real-time qPCR
• TP53

Introduction

Ovarian cancer is the most lethal gynecological malignancy. In the last decade, taxanes combined with cisplatin or its analogs (TP regimen) have been considered as a standard first-line treatment for ovarian cancer.^1,2 This therapy replaced platinum–cyclophosphamide (PC) or other protocols based on DNA-damaging agents (PAC, platinum–doxorubicin–cyclophosphamide, or monotherapy). Although the introduction of taxanes significantly improved treatment results, the overall survival rates are still poor in patients with advanced disease.^2-4 Analysis of molecular prognostic factors may reveal novel potential targets of molecular therapy.

Inhibitors of cyclin-dependent kinases (INK4) play a key role in the cell cycle arrest under cellular stress caused, for example, by chemotherapeutic drugs. The INK4 family includes four members, p16^INK4a, p15^INK4b, p18^INK4c, and p19^INK4d, that specifically bind to and inhibit the kinase activity of CDK4/6 by preventing their interaction with cyclin D. The CDK4/6-cyclin D complex phosphorylates the retinoblastoma protein (pRB), thereby enabling the cell to enter S phase of the cell cycle.⁵ One member of INK4 family, p16^INK4a, is homozygously deleted at a high frequency in a wide variety of human tumor-derived cell lines and several specific types of primary tumors.⁶ Inactivation of other INK4 family genes (p15^INK4b or p18^INK4c) was observed in some cell lines, whereas there is no evidence of their inactivation in solid tumors; even less is known on the role of p19^INK4d protein in tumorigenesis.⁷

p19^INK4d was identified in yeast in 1995.⁸ This 166 amino acid protein is encoded by the CDKN2D (cyclin-dependent kinase inhibitor 2D) gene located on chromosome 19p13. p19^INK4d is highly similar to other INK4 family members, but in contrast to p16^INK4a, it has five ankyrin-like repeats that are thought to mediate protein–protein interactions.⁹ Data obtained from mouse NIH 3T3 cells showed that p19^INK4d protein, like other INK4s, inhibits CDK4–cyclin D1 activity in vivo, and induces G₁ phase arrest.⁸ Recent discoveries suggest that p19^INK4d also plays an important role in the DNA damage response pathway dependent on ATM/ATR kinases. After genotoxic stress caused by agents such as cisplatin, p19^INK4d enters the nucleus.¹⁰ Interestingly, p19^INK4d activities in DNA repair and cell cycle arrest seem to be independent. The protein with mutated phosphorylation sites is unable to enter the nucleus but can still stop cell cycle progression. These data suggest that p19^INK4d status may interfere with cellular responses to cytotoxic agents.

p19^INK4d is expressed ubiquitously in proliferating cultured cells and in normal mouse tissues.^8,11,12 Its expression shows variations during organ development and cell cycle progression.¹³ p19^INK4d expression in cancers has been examined in only a few studies and, to date, it has not been linked to cancer development.^14-16

TP53 accumulation is one of the most frequently observed aberrations in ovarian carcinomas.¹⁷ TP53 dysfunction may enhance response of ovarian cancer to taxane–platinum treatment.^4,18,19 Regarding cisplatin, impaired TP53 protein function may contribute to resistance to this drug, as observed in cell lines and some clinical studies.^20-23 The results obtained in recent years suggest that the TP53 status may influence the clinical importance of other molecular factors.^24-26

We aimed to evaluate the prognostic and predictive value of p19^INK4d expression at the protein and mRNA level with respect to TP53 status in a large group of ovarian cancer patients treated with two different chemotherapy modules. We also searched for CDKN2D gene alterations in these tumors.

Results

Analysis of p19^INK4d protein expression

p19^INK4d expression was observed in both the nuclei and the cytoplasm of ovarian cancer cells, but predominantly in the former (Fig. 1). High p19^INK4d expression (overall score ≥4) was found in 65% (288 of 446) of tumors. Rates for tumors with score category 0, 1, 2, 3, 4, 5, and 6 were 0%, 2%, 7%, 27%, 41%, 23%, and 1%, respectively (Table 1).

Figure 1. p19^INK4d protein expression in two different ovarian carcinomas (400× , hematoxylin counterstain): (A) low p19^INK4d expression, (B) high p19^INK4d expression.

Table 1. p19^INK4d and TP53 expression in ovarian carcinomas

	All patients n = 446	TP-treated group n = 199	PC/PAC-treated group n = 246
p19^INK4d expression:
Low (overall score ≤ 3)	158 (35%)	71 (36%)	87 (35%)
High (overall score ≥ 4)	288 (65%)	128 (64%)	160 (65%)
TP53 expression:
TP53-positive carcinomas (≥ 10%)	263 (59%)	118 (59%)	145 (59%)
TP53-negative carcinomas (< 10%)	182 (41%)	81 (41%)	101 (41%)

The increased p19^INK4d immunostaining was detected in 218 serous (218 of 342, 69%), 16 endometrioid (16 of 22; 73%), 8 clear cell (8 of 15, 53%), 25 undifferentiated (25 of 33, 76%), and in 18 tumors of other histological types (18 of 31, 58%). There was no association between p19^INK4d expression and tumor histology. High p19^INK4d expression was more common in poorly differentiated (86 of 120, 72%) than in moderately and well differentiated carcinomas (172 of 269, 64%; 29 of 56, 52%, respectively; P = 0.035). It was also generally more frequent in higher clinical stages than in lower ones (P = 0.037 in linear trend). High expression of p19^INK4d was found in 39 of 50 (78%) carcinomas in FIGO stage IV, 185 of 284 (65%) in stage IIIC, 45 of 84 (54%) in stages III A, B; unexpectedly, p19^INK4d expression was also high in FIGO stages IIB and C—in 18 of 27 (67%) tumors. High p19^INK4d expression was significantly more frequent in cases with residual tumor size (RT) >2 cm (148 of 211, 70%) than in those with RT 0.5–2 cm (89 of 145, 61%) or complete debulking (50 of 89, 56%) (P = 0.044).

We investigated the potential relationship between the intensity of p19^INK4d immunostaining and patients outcomes with respect to TP53 accumulation status.

p19^INK4d expression in the platinum/cyclophosphamide-treated group

In the group of patients treated with PC/PAC regimen (n = 246) there was an association between high p19^INK4d expression and complete remission of the disease (CR). Patients who reached CR had lower frequency of high p19^INK4d expression (78 of 143; 55% patients) than those who had other responses (81 of 112; 72% patients). In the univariate analysis, high p19^INK4d expression negatively influenced the probability of complete remission (OR 0.53, P = 0.022; Table 2). This result was not confirmed by the multivariate analysis.

Table 2. Associations of p19^INK4d expression with clinical endpoints in ovarian cancer* (univariate Cox proportional hazards and logistic regression models)

	All patients		TP53^− group
	OR/HR [95% CI]	P value	OR/HR [95% CI]	P value
Platinum-treated group
Complete remission
p19^INK4d expression
Low (overall score ≤ 3)	1.0		-
High (overall score ≥ 4)	0.53 [0.31, 0.91]	0.022	-
Overall survival
p19^INK4d expression
Low (overall score ≤ 3)	-		1.0
High (overall score ≥ 4)	-		1.59 [1.02, 2.46]	0.035**
Taxane–platinum-treated group
Overall survival
p19^INK4d expression
Low (overall score ≤ 3)	-		1.0
High (overall score ≥ 4)	-		2.03 [1.15, 3.58]	0.015**

* There were no associations with p19^INK4d in the TP53⁺ group. **Kaplan–Meier curve

In the subgroup of patients without TP53 accumulation in their cancers (n = 101), median overall survival time was significantly shorter in case of high p19^INK4d protein expression than in case of low level of p19^INK4d (20 vs. 29 mo, P = 0.035). In both the univariate and multivariate analysis, high p19^INK4d expression increased the risk of death (HR 1.59, P = 0.035, Fig. 2A; Table 2; HR 1.61, P = 0.049, Table 3, respectively). Overall survival also showed an association with FIGO stage and debulking status (Table 3). Disease-free survival, complete remission and platinum sensitivity did not associate with p19^INK4d expression in the TP53⁻ subgroup.

Figure 2. Prognostic significance of p19^INK4d (CDKN2D) expression. Patients treated with cisplatin/cyclophosphamide (A) and taxane–platinum (B) regimens with high p19^INK4d expression had a significantly worse overall survival than patients with low p19^INK4d expression (Kaplan–Meier curve). CDKN2D gene expression at the mRNA level correlated with enhanced risk of death in taxane–platinum-treated group (C).

Table 3. Associations of p19^INK4d expression and clinico-pathological factors with overall survival in ovarian cancer patients* (multivariate Cox proportional hazards models)

	TP53^− group
	HR [95% CI]	P value
Platinum-treated group
p19^INK4d expression
Low (overall score ≤ 3)	1.0
High (overall score ≥ 4)	1.61 [1.00, 2.58]	0.049
Residual tumor size
0	1.0
≤ 2 cm	-
> 2 cm	1.63 [1.03, 2.58]	0.034
FIGO stage
IIB, IIC	0.12 [0.02, 0.53]	0.005
IIIA, IIIB	-
IIIC	1.0
IV	-
Taxane–platinum-treated group
p19^INK4d expression
Low (overall score ≤ 3)	1.0
High (overall score ≥ 4)	2.00 [1.13, 3.54]	0.016

* There were no associations with p19^INK4d expression in the entire group and TP53⁺ subgroup.

In the TP53⁺ subgroup (n = 145), there were no associations between clinical endpoints and p19^INK4d protein expression.

p19^INK4d expression in the taxane/platinum-treated group

In the entire group of patients treated with the TP regimen (n = 199) and in the TP53⁺ subgroup (n = 110), there were no associations between clinical endpoints and the expression of the p19^INK4d protein.

Similarly to the above-described findings for the PC/PAC-treated group, in the subgroup of patients without TP53 accumulation in their tumors (n = 81), median overall survival was significantly shorter for patients with high level of p19^INK4d protein than for those with low level of p19^INK4d (27 vs. 42.5 mo, P = 0.015). In both the univariate and multivariate analysis, high p19^INK4d expression increased the risk of death (HR 2.03, P = 0.015, Fig. 2B; Table 2; HR 2.00, P = 0.015, Table 3, respectively). Disease-free survival, complete remission and platinum sensitivity did not associate with p19^INK4d expression.

Analysis of CDKN2D gene expression

We examined whether there was an association between high CDKN2D mRNA level and patients’ outcome.

In accordance with the IHC results, CDKN2D mRNA expression showed an association with overall survival in the TP53⁻ subgroup of patients treated with TP regimen (n = 29). High CDKN2D mRNA levels associated with an enhanced risk of death in both the univariate and multivariate analysis (HR 2.94, P = 0.008, Fig. 2C; HR 4.24, P = 0.001, Table 4, respectively).

Table 4. Associations of CDKN2D mRNA expression and clinico-pathological factors with overall survival in taxane-platinum-treated group* (uni- and multivariate Cox proportional hazards model)

	TP53^− group
	HR [95% CI]	P value
Univariate analysis
CDKN2D mRNA expression	2.94 [1.33, 6.49]	0.008
Multivariate analysis
CDKN2D mRNA expression	4.24 [1.80, 9.98]	0.001
Histological grade
G2	1.0
G3	-
G4	6.11 [2.00, 18.70]	0.001

* There were no associations with CDKN2D mRNA expression in the entire group and TP53⁺ group.

In the entire patients group (n = 74) and in the TP53⁺ subgroup (n = 45), as well as in the group of patients treated with the PC/PAC regimen (only 11 TP53-negative carcinomas), there were no associations between CDKN2D gene expression and clinical endpoints.

Analysis of gene mutations and polymorphisms

We found no gene alterations except for one synonymous substitution. The c.90C>G (p.R30R; rs1968445) polymorphism in exon 1 of the CDKN2D gene was present in seven of 68 cases (10.3%).

Discussion

This study has revealed a new prognostic factor in solid tumors, i.e., the p19^INK4d from the family of genes encoding cyclin-dependent kinase inhibitors. Its expression at the mRNA and/or protein level was found to negatively influence overall survival in a large group of ovarian cancer patients. This association was clearly demonstrated in patients treated either with platinum–cyclophosphamide (PC/PAC) or taxane–platinum (TP) regimens, but only in the subgroup without TP53 accumulation in ovarian carcinomas.

In the literature there are few data on p19^INK4d expression in human tumors and its role in tumorigenesis. p19^INK4d expression in paraffin-embedded tissue was first analyzed by Thullberg et al.²⁷ in 2000. In agreement with our results, they observed the p19^INK4d protein in the nucleus and cytoplasm. It is noteworthy that the mouse monoclonal antibody (clone DCS-100), used in that and the current study, does not give any cross-reactions with other cellular proteins present in whole-cell lysates of human cell lines. Bartkova et al.¹⁴ also observed p19^INK4d expression in both nuclei and cytoplasm in normal human testis, but not in testicular cancer. However, there are some studies on cell lines demonstrating that p19^INK4d has only a cytoplasmic location.^8,11,28 Ceruti et al.²⁸ observed in human neuroblastoma cell line SH-SY5Y that p19^INK4d protein translocated into the nucleus after UV irradiation. It appears that the translocation is not a constitutive process but occurs in response to cell cycle regulatory signals and external stimuli such as UV or cisplatin treatment.^10,28

The results obtained by our group show that high p19^INK4d protein expression associates with adverse clinicopathological factors, i.e., high tumor grade, advanced FIGO stage and bulky residual tumor. Despite the aggressive tumor features, high p19^INK4d protein expression seems to be an independent negative prognostic factor, but this was not consistent in all patients evaluated.

There are two other analyses of clinical importance of p19^INK4d immunohistochemical expression in cancer patients: Bartkova et al.^14,15 studied 71 testicular cancers but these tumors did not express the p19^INK4d protein. Zhang et al.¹⁶ did not find any prognostic value of p19^INK4d protein expression (and also CDKN2D mRNA expression) in patients with nasopharyngeal carcinoma. Thus, the current study and that of Zhang et al.¹⁶ are the only ones evaluating CDKN2D mRNA expression in cancer cells and its impact on clinical endpoints. The negative results of these two analyses are in agreement with each other and generally with the lack of clinical importance of p19^INK4d protein expression in patients studied.

Our analysis revealed the negative impact of p19^INK4d expression (at the protein and mRNA level) only in the subgroup with TP53-negative carcinomas. In the majority of tumors, TP53 immunohistochemical accumulation reflects a TP53 gene missense mutation and TP53 dysfunction, while the lack of TP53 accumulation reflects preserved TP53 function or loss of function due to a nonsense mutation. As previously shown for several proteins, the TP53 immunohistochemical status may determine their clinical significance, particularly if they are regulated by, or interfere with TP53 during the control of tumor cell proliferation or apoptosis.^24-26,29

In the literature there are not yet any data connecting the TP53 protein to the p19^INK4d protein; nevertheless, they participate in the regulation of the same processes, i.e., proliferation, apoptosis and DNA repair. p19^INK4d is activated in response to stress caused by some genotoxic agents such as cisplatin.³⁰ Some experimental results indicate that the p19^INK4d overexpression enhances DNA repair and cell viability.^30,31 In cell lines deprived of p19^INK4d, repair of DNA lesions was impaired and apoptosis was enhanced.^30-32 In a previous study, p19^INK4d-deficient mice did not develop cancers, and apoptosis was found to be increased.³³ Therefore, low levels of p19^INK4d protein expression may decrease the efficiency of DNA repair processes, which may result in a higher level of lesions and the activation of apoptosis. Some studies indicate that p19^INK4d may promote cell viability: the expression of p19^INK4d enhances cell survival after UV treatment, while silencing of its expression increases sensitivity to genotoxic agents in comparison with cells overexpressing p19^INK4d.³¹ Our results suggest that the effect of p19^INK4d action may be particularly efficient in cells with functional TP53, which is when it becomes clinically visible. Further studies are necessary to elucidate whether there is any interaction between p19^INK4d and TP53.

Mutations in CDKN2D gene were also described in various tumor types. The analysis of 67 patients with osteosarcomas showed alterations of the CDKN2D gene, including one frameshift mutation, and five gene rearrangements.³⁴ A frameshift mutation in exon 1 (10%) was also reported in two ovarian cell lines but in no other sample in a large series of cell lines and tumors.^35,36 In our study we did not observe any deletions or point mutations in the CDKN2D gene in a group of 68 patients with ovarian cancer. Only one genetic polymorphism at codon 30 (c.90C>G) in exon 1 of the CDKN2D gene was identified in 10% of samples. Other studies did not report the presence of that alteration in a variety of neoplasms, including hematopoietic neoplasms, cervical, breast and prostate cancer.^34,37,38

In conclusion, our results show for the first time that high p19INK4d expression associates with aggressive clinicopathological cancer features and enhances the risk of death in patients with TP53-negative ovarian carcinomas. This confirms our previous observations that analyses of uniform groups, such as with regard to the TP53 accumulation status, facilitates the identification of cancer biomarkers.

Materials and Methods

Patients and tumors

The study was performed on 445 archival samples of ovarian carcinomas taken during staging laparotomy, before administration of chemotherapy. Medical records were critically reviewed by at least two clinicians. The patients were treated with standard PC (cisplatin−cyclophosphamide or carboplatin−cyclophosphamide) or PAC chemotherapy (PC and doxorubicin) (246 patients), or with taxane–platinum chemotherapy (TP: paclitaxel or docetaxel with cisplatin or carboplatin) (199 patients).⁴ The material was carefully selected out of a total of 899 cases submitted to meet additional criteria: adequate staging procedure, International Federation of Gynaecologists and Obstetricians (FIGO) stage IIB to IV disease.³⁹ tumor tissue from the first laparotomy available, moderate (G2) or poor tumor differentiation (G3 and G4), availability of clinical data, including residual tumor size and follow-up.

All tumors were uniformly reviewed histopathologically, classified according to the criteria of the World Health Organization⁴⁰ and graded in a four-grade scale, according to the criteria given by Broders.⁴¹ Clinicopathological characteristics are presented in Table 5.

Table 5. Patients’ characteristics

	All patients n = 445	TP-treated group n = 199	PC/PAC-treated group n = 246
Age (years)
Range	20–78	20–78	24–77
Mean	54.3	54.9	53.9
FIGO stage
IIB, IIC	27 (6%)	10 (5%)	17 (7%)
IIIA, IIIB	84 (19%)	28 (14%)	56 (23%)
IIIC	284 (64%)	141 (71%)	143 (58%)
IV	50 (11%)	20 (10%)	30 (12%)
Residual tumor size
0	89 (20%)	37 (18%)	52 (21%)
≤ 2 cm	144 (32%)	80 (40%)	64 (26%)
> 2 cm	212 (48%)	82 (42%)	130 (53%)
Histological type
Serous	342 (77%)	148 (74%)	194 (79%)
Endometrioid	22 (5%)	8 (4%)	14 (6%)
Clear cell	15 (3%)	4 (2%)	11 (4%)
Undifferentiated	34 (8%)	21 (11%)	13 (5%)
Other types	32 (7%)	18 (9%)	14 (6%)
Histological grade
G 2	56 (13%)	26 (12%)	30 (12%)
G 3	268 (60%)	114 (58%)	154 (63%)
G 4	121 (27%)	59 (30%)	62 (25%)
Response to chemotherapy
Complete remission	265 (59%)	131 (65%)	134 (55%)
Partial remission/no change	113 (26%)	62 (32%)	51 (20%)
Progression	67 (15%)	6 (3%)	61 (25%)
Platinum-sensitive	133 (30%)	68 (34%)	65 (26%)
Platinum-resistant	225 (51%)	87 (45%)	138 (56%)
Number of patients at risk (OS)
1 y	398 (89%)	188 (94%)	210 (85%)
2 y	284 (56%)	148 (74%)	136 (55%)
3 y	187 (42%)	98 (49%)	89 (36%)
4 y	124 (28%)	61 (31%)	63 (26%)
5 y	87 (20%)	42 (21%)	45 (18%)
Number of patients at risk (DFS)
1 y	150 (57%)	76 (58%)	74 (55%)
2 y	87 (33%)	44 (34%)	43 (32%)
3 y	65 (25%)	32 (24%)	33 (25%)
4 y	49 (19%)	23 (18%)	26 (19%)
5 y	32 (12%)	14 (11%)	18 (13%)
Follow-up time
Range (month)	4.4–198.3	4.8–103.4	4.4–198.3
Mean	30	35.5	27.5
Outcome
NED	46 (10%)	28 (14%)	18 (7%)
AWD	47 (10%)	36 (18%)	11 (4%)
DOD	342 (77%)	130 (65%)	212 (87%)
DOC	10 (3%)	5 (3%)	5 (2%)

PC, cyclophosphamide and cisplatin; PAC, PC plus doxorubicin; TP, taxane–platinum therapy; NED, no evidence of disease; AWD, alive with disease; DOD, died of disease; DOC, died of other causes; OS, overall survival; DFS, disease free survival.

For the PC/PAC-treated group, the follow-up time ranged from 4.4 to 198.3 mo (median: 27.5); the respective values for the TP-treated group were: 4.8 to 103.4 mo (median: 35.5). Short follow-up times resulted from early patient deaths. All surviving patients had at least a 6 mo of follow-up. Response to chemotherapy was evaluated retrospectively according to the World Health Organization response evaluation criteria.⁴² The evaluation was based on data from medical records describing the patient’s clinical condition and CA125 levels in 3–4 week intervals. Complete remission (CR) was defined as the disappearance of all clinical and biochemical symptoms of ovarian cancer evaluated after completion of first-line chemotherapy and confirmed at four weeks. Within the CR group, we identified a platinum-sensitive group (PS, disease-free survival longer than six months). Other tumors were described as platinum-resistant.⁴³

The study was approved by the bioethics committee of the Institute of Oncology (39/2007).

Immunohistochemical analysis

Immunohistochemical stainings were performed on paraffin-embedded material after heat-induced epitope retrieval (HIER). We used a mouse monoclonal anti-p19^INK4d antibody (1/175, Sigma-Aldrich). TP53 protein was detected with the use of PAb1801 monoclonal antibody (1/3000, Sigma-Genosys), as described previously.²⁴ The antigens were retrieved by heating the sections in 0.01 M citrate buffer (pH 6.0): 2 h at 99 °C in a water bath for p19^INK4d and 2 × 5 min for TP53 at 700 W in a microwave oven. Non-specific tissue and endogenous peroxidase reactivities were blocked with 10% BSA and 3% H₂O₂, respectively. The sections were incubated with primary antibodies overnight, at 4 °C. Biotinylated secondary goat anti-mouse IgG (1/1500), peroxidase-conjugated streptavidin (1/500) (all from Immunotech), and DAB were used as a detection system. As a positive control for TP53 accumulation, we used a tumor with a defined TP53 gene missense mutation.⁴⁴ Normal mouse IgG of the same subclasses and at the concentrations of the relevant primary antibodies served as negative controls.

The evaluation of immunohistochemical stainings

p19^INK4d expression was scored independently for nuclear and cytoplasmic staining. Light microscopic evaluation at 400× magnification was used to count tumor cells within the areas of the strongest staining. p19^INK4d expression was categorized according to the staining intensity (0 or weak to strong: 1 to 3). An overall protein expression score was calculated by adding the nuclear and cytoplasmic scores (overall score range, 0–6). This overall score for each case was further simplified by dichotomizing it to low (overall score of ≤3) or high (score of ≥4). Two independent assessors (AFG and AR) concurred in 80% of the cases, and reached consensus in the remaining cases. TP53 protein accumulation was assessed as present (more than 10% positive cells) or absent, as previously described.⁴

CDKN2D expression at the mRNA level

Real-time PCR experiments were conducted on 74 of 199 ovarian cancer samples from patients treated with a taxane–platinum regimen and on 32 of 426 samples from patients treated with a platinum/cyclophosphamide regimen. Only tumors containing less than 5% of non-tumor tissue were analyzed at mRNA level. Total RNA was extracted from frozen tissues with the Nucleo Spin RNA kit (Mecherey Nagel) according to the manufacturer’s protocol. RNA quantity was measured by UV spectrophotometry, and its quality was assessed by the 260/280 ratio and in a 1% agarose gel. One microgram of total RNA was transcribed to cDNA using the Super Script III First Strand kit (Invitrogen,). Quantitative RT-PCR (Q-PCR) was run on the 7500 Fast Real-Time PCR System (Applied Biosystems), with the use of the TaqMan Gene Expression Assays (Hs00176481_m1 for CDKN2D and 4326321E for the reference gene HGPRT). All real-time PCR reactions were performed in triplicate in the final volume 10 μl for 40 cycles according to the following protocol: each cycle at 95 °C for the initial 10 min, then at 95 °C for 15 s, and 60 °C for 1 min. The obtained results were averaged, and gene expression levels were normalized to the HGPRT expression. A standard curve, used in all experiments, was prepared from serial dilutions of cDNA from one of the tumors analyzed.

Analysis of the CDKN2D gene mutations and polymorphisms

The analysis of genetic alterations of the CDKN2D gene was performed for 68 ovarian tumors. Genomic DNA was isolated from frozen tissues using the QIAamp DNA Extraction Kit (Qiagen) following the kit’s instructions. Molecular analyses were performed by the PCR-SSCP and sequencing methods. The first exon of CDKN2D was amplified with primers (1F) 5′-TTTGCAGGCC GCCAGTGTC-3′, and (1R) 5′-GTCTCGATCC TCATCCCGCT TAGCC-3′. The second exon was analyzed as two fragments with primers 2a: (2a-F) 5′-CTGATCCTCT GTCCCTCACA C-3′ and (2a-R) 5′-TGAACTGCCA GATGGATTGG AAGT-3′, and 2b: (2b-F) 5′-ATTCCTGGAC ACCCTGAAG GTCCT-3′, and (2b-R) 5′-GGGCAGGAGA AACAAGAAGA GAAAG-3′. The CDKN2D DNA sequence accession number AF518878.1 (GenBank) was used to design the primers sequences. The PCR mixture was prepared according to the protocol provided with AmpliTaq Gold PCR kit (Life Technologies) with the addition of 5 mM betaine (Sigma-Aldrich). Thermal cycling conditions were as follows: an initial denaturation at 94 °C for 10 min; followed by 36 cycles of denaturing at 94 °C for 30 s, annealing at 60 °C (exon 1 and fragment 2b) or 55 °C (fragment 2a) for 30 s, and extension at 72 °C for 30 s, with a final 7 min extension at 72 °C. Electrophoresis for SSCP analysis was performed as previously described.⁴⁵ DNA from the samples with bands shifts detected by SSCP were sequenced with ABI dye terminator sequencing kit according to the manufacturer's recommendations in an ABI PRISM 3100 sequencer (Life Technologies).

Statistical analysis

The associations between p19^INK4d expression and clinicopathological variables were assessed using the chi-square test. Overall survival (OS) and disease-free survival (DFS) were estimated using the Kaplan–Meier method and a log-rank test. Overall and disease-free survival time analyses were performed using multivariate Cox proportional hazard models. Tumor response to chemotherapy (probability of CR, PS, or HPS) was evaluated using a multivariate logistic regression model.

Statistical analyses included the following independent variables: age of patients (median: 53 y), the FIGO stage, histopathological type, grade, residual tumor size and TP53 accumulation status. The p19^INK4d expression was analyzed as a categorical variable (cut-off point at median: 3). Important factors were selected using a backward selection technique, where factors not significant at 0.1 were removed stepwise from the model.

The analyses were performed in the two groups of patients treated with different chemotherapeutic regimens, and additionally in the TP53^– and TP53⁺ subgroups. All tests were two-sided. A P value < 0.05 was considered significant. All calculations were performed using STATA 5 software.

Abbreviations:
CR	=	complete remission
DFS	=	disease-free survival
HR	=	hazard ratio
OR	=	odds ratio
OS	=	overall survival
PC/PAC	=	platinum–cyclophosphamide chemotherapy
PS	=	platinum sensitivity
TP	=	taxane-platinum chemotherapy

Acknowledgments

This study was supported by grants 2 PO5A 068 27 and N N401 236134 of the Polish Ministry of Science and Higher education.

10.4161/cbt.25966/

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.