This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Pruneri, G. Articles by Neri, A. Search for Related Content PubMed PubMed Citation Articles by Pruneri, G. Articles by Neri, A. Social Bookmarking                            What's this?

Clinical Relevance of Expression of the CIP/KIP Cell-Cycle Inhibitors p21 and p27 in Laryngeal Cancer

From the II Servizio di Anatomia Patologica, Clinica Otorinolaringoiatrica I, Laboratorio Epidemiologico, Laboratorio di Ematologia Sperimentale e Genetica Molecolare, Servizio di Ematologia, Istituto di Scienze Mediche, Università di Milano, Ospedale Maggiore di Milano, Istituto di Ricovero e Cura a Carattere Scientifico (IRCCS), Milan, Italy.

Address reprint requests to Antonino Neri, MD, Servizio Ematologia, Centro "G. Marcora," Ospedale Maggiore, IRCCS, Via F. Sforza 35, 20122, Milano, Italy; email sap2{at}imiucca.csi.unimi.it

	ABSTRACT

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
PURPOSE: To investigate the prognostic relevance of p21 and p27 protein expression in laryngeal squamous cell carcinoma (LSCC).

PATIENTS AND METHODS: We have analyzed by immunohistochemistry p21 and p27 expression in a series of 132 patients who underwent surgical resection of their LSCC and who had previously been investigated for p53 gene mutations and cyclin D1 expression. The tumors were considered low expressors when they had 10% of p21 and 50% of p27 immunoreactive neoplastic cells.

RESULTS: In 41 cases (31.1%), p21 was expressed in 10% of neoplastic cells; in 91 cases (68.9%), it was expressed in more than 10% of neoplastic cells. In 11 cases (8.3%), p27 was expressed in less than 5% of neoplastic cells; in 39 cases (29.6%), it was expressed in 5% to 50% of neoplastic cells; and in 82 cases (62.1%), it was expressed in more than 50% of the neoplastic cells. Low levels of p21 expression were associated with poor histologic differentiation and lymph node metastases. Low levels of p27 expression were associated with tumor extension and advanced clinical stage. Expression of p21 and p27 was not correlated with p53 gene status, and low p27 expression was more frequently detected in the cyclin D1–positive cases, with a borderline level of statistical significance. At univariate analysis, anatomic site, tumor extension, clinical stage, high cyclin D1 expression, and low p27 expression were significantly associated with reduced disease-free and overall survival rates. At multivariate analysis, high cyclin D1 expression and low p27 expression were the only significant covariates. The patients with a cyclin D1⁺/p27^- phenotype had the poorest disease-free and overall survival rates.

CONCLUSION: Our study provides evidence that the immunohistochemical evaluation of p27 expression is a significant independent predictor of prognosis in laryngeal carcinoma.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
TUMOR CELLS TYPICALLY show acquired damage to the genes involved in cell cycle control. Progression from the G1 to S phase in the mammalian cell cycle is closely regulated by the formation of cyclin/cyclin-dependent kinase (cdk) complexes.¹ The kinase activity of cdks is inhibited by a number of specific proteins belonging to the INK4 and CIP/KIP families.² The CIP/KIP family consists of p21 (WAF1/Cip1),³ p27 (Kip1),⁴ and p57 (Kip2).⁵ p21 and p27 can induce G1 arrest by inhibiting cyclin D–, E–, and A–dependent kinases.^3,4 p21 is a downstream effector of p53-induced cell-cycle arrest⁶ and apoptosis⁷ following DNA damage, although there is evidence that other mechanisms induce p21 expression independently of p53.^8,9 The intracellular levels of p27 increase in response to contact inhibition as well as to a number of extracellular antimitogenic signals.¹⁰

Unlike other inhibitors of cell-cycle progression, genetic alterations of p21 and p27 are uncommon,^11,12 but experimental observations have demonstrated that their levels in human tumors may vary as a result of posttranscriptional regulation.^13,14 Moreover, it has been suggested that low p27 levels may be due to increased proteasome-mediated degradation activity rather than altered gene expression in non–small-cell lung¹⁵ and colorectal carcinomas.¹⁶ Interestingly, the loss of p21 and p27 expression has been found to be associated with a poor prognosis in a number of different human tumors, including non–small-cell lung,^15,17 breast,^18,19 prostate,^20,21 and gastric^22,23 carcinomas. However, the clinical relevance of p21 and p27 expression in head and neck cancer has not yet been fully elucidated. p21 expression was found to have no prognostic significance in two different studies based on carcinomas of the pyriform sinus²⁴ and larynx.²⁵ On the other hand, it has been reported that high levels of p21 correlate with an adverse clinical course in a small series of tumors of the head and neck region.²⁶ To the best of our knowledge, the prognostic significance of p27 expression in head and neck tumors has not been investigated so far.

To improve our understanding of the role and clinical relevance of p21 and p27 in laryngeal cancer, we have studied, using immunohistochemistry, their expression in a homogeneous series of 132 patients who had previously been investigated for alterations of other genes involved in cell-cycle regulation, such as p53²⁷ and cyclin D1.²⁸

	PATIENTS AND METHODS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
One hundred forty-nine consecutive patients with laryngeal squamous cell carcinoma (LSCC) were selected from the 1983 to 1995 files of Clinica Otorinolaringoiatrica I of Milan's School of Medicine. The patients had to fulfill the following inclusion criteria: no history of previous malignancies, primary squamous cell carcinoma of the larynx only, no previous radiotherapy or chemotherapy, and surgical excision of the tumor by means of partial or total laryngectomy. The patients whose lymph nodes were proved to be metastatic by means of conventional histology underwent postsurgical adjuvant locoregional radiotherapy (50 to 60 Gy). Seventeen of the 149 selected patients were excluded from the study because no tumor samples were available for immunohistochemical analysis. Forty of the remaining 132 patients had previously been investigated for p53 gene mutations²⁷ and all of them were examined for cyclin D1 expression.²⁸ Written informed consent was obtained from each patient included in the study.

The main clinical and pathologic characteristics of the patients are shown in Table 1: 129 (97.7%) were male and three (2.3%) were female; their mean age was 61.5 years (SD, 10.2; range, 23 to 90 years; median age, 62 years). All of the patients had been exposed to risk factors such as tobacco smoke and/or alcohol consumption. The information regarding smoking included smoking status (never or current smoker), the duration of exposure, and the number of pack-years of smoking. The duration of exposure ranged from 15 to 60 years (mean, 30.3 years; median, 30 years; SD, 11.2 years); the patients were grouped into three classes on the basis of the number of pack-years of smoking (0 to 19 pack-years, 20 to 45 pack-years, and > 45 pack-years). Forty-nine patients (37.1%) had smoked for 0 to 19 pack-years, 42 patients (31.8%) for 20 to 45 pack-years, and 41 patients (31.1%) for more than 45 pack-years. The information regarding alcohol consumption included the number of drinks per day before diagnosis (one drink corresponding to 150 mL of wine or 30 mL of hard liquor). Sixty-one patients (46.2%) were classified as heavy drinkers ( seven drinks per day), and 71 (53.8%) were considered moderate (< seven drinks per day)/nondrinkers.

The clinical staging and identification of the anatomic site of the tumors were based on the International Union Against Cancer's tumor-node-metastasis staging system.²⁹ Seventy-three tumors (55.3%) were supraglottic and 59 (44.7%) were glottic. Thirty-nine tumors (29.6%) were in stage I, 35 (26.6%) were in stage II, 29 (21.9%) were in stage III, and 29 (21.9%) were in stage IV. Diagnosis and assessment of the grade of differentiation of the tumors according to the method of Shanmugaratnam et al³⁰ showed that there were 23 (17.4%) well (G1), 69 (52.3%) moderately (G2), and 40 (30.3%) poorly (G3) differentiated squamous cell carcinomas. The follow-up of the patients ranged from 4 to 181 months (mean follow-up, 75.8 months; median, 74.0 months; SD, 48.2 months). During the follow-up period, 43 tumors (32.6%) relapsed. Forty patients (30.3%) died because of the disease, and eight (6.1%, who were disease-free) died of other causes; of the 84 surviving patients (63.6%), 81 were free of disease, and three were alive with disease.

Immunohistochemistry
One hundred thirty-two surgical specimens of LSCC were analyzed by means of immunohistochemistry; no bioptic samples of carcinoma were included in the study. In each case, we analyzed a single section considered to be representative of the tumor in terms of the pattern of infiltration, histologic differentiation, lymphoid infiltrate, and vascular invasion. Moreover, 50 samples of normal mucosa were also analyzed (40 surgical specimens of normal mucosa adjacent to infiltrative carcinoma and 10 bioptic samples without any histologic signs of dysplasia taken from patients who had never developed an infiltrating tumor). Finally, we studied p21 and p27 expression in 13 lymph node metastases (10 massive and three submassive) taken from 10 cases (seven pN1, two pN2b, and one pN2c) in which the corresponding primary tumors had the following phenotypes: p21^-/p27^- (two cases), p21^-/p27⁺ (three cases), p21⁺/p27⁺ (one case), and p21⁺/p27^- (four cases). All of the samples were routinely fixed in 10% buffered formalin and embedded in paraffin. p21 and p27 protein expression was assayed by means of the avidin-biotin peroxidase complex method, using the 3,3'-diaminobenzidine tetrahydrochloride chromogen as reported.²⁸ For antigen retrieval, the slides were placed in a 0.01 mol/L citrate buffer at pH 6.0 (for p21) or in a 0.01 mol/L EDTA buffer at pH8.0 (for p27) and underwent three (for p21) and four (for p27) 5-minute 780 W cycles at 90°C in a microwave oven. The sections were immunostained with the anti-p21 EA10 (Oncogene Science Inc, Cambridge, MA) and anti-p27 1B4 (Novocastra Laboratories Ltd, Newcastle upon Tyne, United Kingdom) monoclonal antibodies. The working concentration was 1:100 for both antibodies. Formalin-fixed and paraffin-embedded samples from colon (for p21) and breast (for p27) carcinomas were used as positive controls; the morphologically normal laryngeal epithelium was used as an internal positive control for both antibodies. Small lymphocytes and plasma cells, which have been reported to express p27,³¹ served as an internal positive control for the anti-p27 antibody. Nonimmune mouse serum, as a substitution for the primary antibody, was used as a negative control for both the antibodies. Hyperplastic germinal centers, which are known to be devoid of p27 protein,³¹ served as an internal negative control for the anti-p27 antibody.

The immunoreactive cells were independently evaluated by two of us (G.P. and N.C.), who did not know the clinicopathologic features and the patients' clinical course. At least 20 high-power fields were analyzed for each case, and all of the immunoreactive cells were considered positive; the percentage of p21 and p27 positive cells was expressed as a ratio of positive cells to the total number of cells counted. In accordance with previous studies,^18,32,33 tumors with 10% of p21 immunoreactive cells were considered low expressors. For p27, we used the scoring system proposed by Esposito et al¹⁵ in non–small-cell lung cancer: less than 5% of positive cells, 5% to 50% of positive cells, and more than 50% of positive cells. The cases with p27 expression in 50% of the neoplastic cells were considered low expressors in accordance with the publishing convention.^19,34 The interobserver reproducibility was 93.2% for p21 and 90.9% for p27.

Statistical Analysis
Descriptive statistics were calculated with their 95% confidence intervals (CIs). The associations between the clinicopathologic features and p21 and p27 expression (low or high expressors) were evaluated by means of the ² test with Yates' correction and the ² test for trend in the case of ordinal variables.³⁵ The prognostic relevance of patient baseline characteristics on disease-free survival and overall mortality (death due to cancer and other causes) was assessed by means of survival analysis based on the Kaplan-Meier product-limit estimate of the cumulative probability survival function.³⁶ The follow-up of the patients was truncated at 132 months (with more than 13% of the starting sample cases still at risk), in order to obtain satisfactorily precise probability estimates. The prognostic relevance of p21 and p27 expression was assessed using two classes for p21 ( 10% and > 10%) and, in accordance with previous studies,^16,19 two ( 50% and > 50%) or three (< 5%, 5% to 50%, and > 50%) classes for p27. The other investigated prognostic factors were classified as follows: age at diagnosis ( 60 years and > 60 years), smoking exposure (0 to 19, 20 to 45, and > 45 pack-years), alcohol consumption ( seven drinks per day and < seven drinks per day/nondrinkers), histologic grade (G1, G2, and G3), lymph node metastasis (presence and absence), tumor extension (T1 to T2 and T3 to T4), and clinical stage (I to II and III to IV). The disease-free and overall survival functions of the different classes of investigated variables were compared using the log-rank test.³⁷ In the case of ordinal variables, in addition to the overall log-rank statistic, the pertinent test for trend was also made. Finally, a multivariate analysis was made according to Cox's model following a backward procedure³⁸; since the three classes of p27 expression had different patterns of disease-free and overall survival, they were coded as two dummy variables in the multivariate model. The goodness of fit and the fulfilment of the proportionality hazard assumption were checked using graphical methods.

	RESULTS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
p21 and p27 Immunoreactivity in Normal Mucosa and LSCC
The mean proportions of p21 and p27 expression in normal mucosa adjacent to infiltrative tumors were, respectively, 18.8% ± 4.7% (95% CI, 17.5% to 20.2%; range, 14% to 30%; median, 18%) and 46.4% ± 8.0% (95% CI, 44.0% to 48.6%; range, 30% to 60%; median, 49%). There was no statistically significant difference in p21 and p27 expression between the mucosa of normal subjects and that adjacent to infiltrative tumors (data not shown).

The patterns of expression of p21 and p27 in laryngeal epithelium were exclusively nuclear. p21 was expressed within the intermediate layers, whereas the basal and superficial layers were usually unreactive (Fig 1A). p27 was expressed throughout the epithelium, with the exception of the most superficial layers (Fig 1B). The anti-p21 and -p27 antibodies variably labeled the nuclei of fibroblasts, smooth muscle fibers, and seromucinous glands and ducts; furthermore, nuclear p27 expression was invariably seen in small lymphocytes and plasma cells, whereas the germinal centers of the mucosa-associated lymphoid tissue were unreactive (Fig 1B and 1E and data not shown).

View larger version (139K): [in this window] [in a new window]   Fig 1. Morphologically normal mucosa showing p21 (A) and p27 (B) nuclear expression. Squamous cell carcinomas with low (C) and high (D) nuclear p21 expression. Mitoses are unreactive (D, arrowheads). Squamous cell carcinomas with nuclear p27 expression in < 5% (E) and > 50% (F) of the cells. The mitosis in panel F is unreactive (arrowhead). p21 (G) and p27 (H) expression in neoplastic infiltrating nodules (x250 original magnification, hematoxylin counterstain).

In 41 cases (31.1%), p21 was expressed in 10% of neoplastic cells (Fig 1C), and in 91 cases (68.9%), it was expressed in more than 10% (Fig 1D). In 11 cases (8.3%), p27 was expressed in less than 5% of neoplastic cells (Fig 1E); in 39 cases (29.6%), in 5% to 50%; and in 82 cases (62.1%), in more than 50% (Fig 1F). As in the case of normal mucosa, the pattern of immunostaining was exclusively nuclear for both antibodies. Mitoses were unreactive (Fig 1D and 1F). p21 expression was more pronounced in the centrally located keratinizing cells of the neoplastic infiltrating nodules (Fig 1G), whereas p27 was homogeneously distributed throughout the tumor (Fig 1H). Complete concordance in p21 and p27 expression was found between the primary tumors and metastatic deposits.

Relationship Between p21 and p27 Expression and p53 Status
We did not find any statistically significant association between p53 gene status and p21 or p27 protein expression (P = .648 and P = .311, respectively) in the 40 samples that had been previously analyzed for p53 gene mutations in exons 5 through 8.²⁷ In particular, high p21 expression was found in eight (61.5%) of the 13 mutated cases and 13 (48.1%) of the 27 cases bearing the wild-type p53 gene; high p27 expression was found in eight (61.5%) of the 13 mutated cases and 12 (44.4%) of the 27 cases bearing the wild-type p53 gene.

Relationship Between p21 and p27 Expression and Cyclin D1 Expression
Cyclin D1 expression (defined as its expression in > 5% of the neoplastic cells) has previously been found to represent a significant independent prognostic factor for both disease-free and overall survival.²⁸

Low levels of p21 expression ( 10% of the neoplastic cells) were found in 15 (33.3%) of the 45 cyclin D1–positive cases (cyclin D1⁺/p21^-) and in 26 (29.8%) of the 87 cyclin D1–negative cases (cyclin D1^-/p21^-) (P = .685). Of the remaining 91 patients, 30 had a cyclin D1⁺/p21⁺ phenotype and 61 had a cyclin D1^-/p21⁺ phenotype.

Finally, low levels of p27 protein ( 50% of the neoplastic cells) were found in 22 (48.9%) of the 45 cyclin D1–positive cases (cyclin D1⁺/p27^-) and in 28 (32.2%) of the 87 cyclin D1–negative cases (cyclin D1^-/p27^-), an inverse correlation that was just above the level of statistical significance (P = .092). Of the 82 remaining patients, 23 had a cyclin D1⁺/p27⁺ phenotype and 59 had a cyclin D1^-/p27⁺ phenotype.

p21, p27, and Clinicopathologic Features
The relationship between the immunohistochemical results and clinicopathologic features is summarized in Table 1: the percentage of p21 low expressors progressively increased from G1 (7.3%) to G2 (39.0%) and G3 (53.7%) tumors (P = .001 and P = .001 for trend); moreover, p21 low expressors were more frequent in node-positive tumors (P = .028). Finally, the proportion of p21 low expressors was higher in cases at an advanced clinical stage and in heavy drinkers (just above the statistical significance threshold, P = .089 and P = .086, respectively).

High levels of p27 expression (> 50% of the neoplastic cells) were significantly more frequent in T1 to T2 tumors (P =.005) and in clinical stages I to II (P = .018). Finally, high levels of p27 expression were more frequent in glottic tumors (just above the statistical significance threshold, P = .080).

Disease-Free Survival
The disease-free survival probability function of the patients at 5 years was 0.69 (95% CI, 0.61 to 0.77). At univariate analysis, anatomic site (P = .0417), tumor extension (P = .0001), clinical stage (P = .0004), cyclin D1 (P = .0002), and p27 expression (P = .0015 comparing two classes and P = .0001 comparing three classes) were significantly associated with tumor recurrence (Table 2).

At multivariate analysis, cyclin D1 (P = .0053) and p27 (P = .0001) were the only statistically significant independent predictors of disease-free survival. In particular, the patients with p27 expression in less than 5% and in 5% to 50% of neoplastic cells had a relative risk of relapse of, respectively, 7.04 (95% CI, 3.12 to 15.90) and 4.55 (95% CI, 1.96 to 10.53) in comparison with those with p27 expression in more than 50% of neoplastic cells. Cyclin D1–positive cases had a relative risk of 2.43 (95% CI, 1.30 to 4.52) in comparison with the negative cases.

Overall Survival
The overall survival probability function of the patients at 5 and 8 years was, respectively, 0.67 (95% CI, 0.59 to 0.79) and 0.62 (95% CI, 0.54 to 0.70). Univariate analysis showed that a shorter overall survival significantly correlated with anatomic site (P = .0422), tumor extension (P = .0042), clinical stage (P = .0072), cyclin D1 overexpression (P = .0064), and low p27 expression (P = .0127 comparing two classes and P = .0001 comparing three classes) (Table 3 and Figs 2 and 3). In particular, the median survival of patients with p27 expression in less than 5%, 5% to 50%, and more than 50% of the cells was 18 months, 66 months, and 93 months, respectively.

View larger version (10K): [in this window] [in a new window]   Fig 2. Overall survival of the patients grouped into two classes of p27 expression ( 50%, 50 patients; > 50%, 82 patients).

View larger version (16K): [in this window] [in a new window]   Fig 3. Overall survival of the patients grouped into three classes of p27 expression (< 5%, 11 patients; 5% to 50%, 39 patients; > 50%, 82 patients).

A Cox proportional hazards model including all of the clinicopathologic variables that had proved to be significant at univariate analysis showed that cyclin D1 overexpression (P = .0494) and low p27 expression (P = .0009) were the only statistically significant covariates. In particular, patients with p27 expression in less than 5% and in 5% to 50% of neoplastic cells had a relative risk of dying of 4.67 (95% CI, 2.08 to 10.48) and 3.39 (95% CI, 1.45 to 7.93), respectively, in comparison with cases with p27 expression in more than 50% of neoplastic cells. Patients with cyclin D1 overexpression had a relative risk of 1.80 (95% CI, 1.00 to 3.24) in comparison with negative cases.

p27 and Cyclin D1 Expression and Survival
Three classes of prognostic clinical relevance for disease-free and overall survival were obtained from the interaction of cyclin D1 and p27 expression in Cox's model with 95% CI of the relative risks (Tables 2 and 3 and Fig 4): cyclin D1^-/p27⁺ with the best prognosis, cyclin D1⁺/p27⁺–cyclin D1^-/p27^- with an intermediate prognosis, and cyclin D1⁺/p27^- with the poorest prognosis (P = .0001 and P = .0001 for trend for disease-free survival; P = .0015 and P = .0008 for trend for overall survival). In particular, the relative risk of tumor recurrence was 5.94 (95% CI, 2.65 to 13.29) in cyclin D1⁺/p27^- patients and 2.74 (95% CI, 1.12 to 5.28) in cyclin D1⁺/p27⁺ or cyclin D1^-/p27^- patients against cyclin D1^-/p27⁺. With regard to overall survival, the relative risk was 3.60 (95% CI, 1.73 to 7.51) in cyclin D1⁺/p27^- patients and 1.72 (95% CI, 0.87 to 3.39) in cyclin D1⁺/p27⁺or cyclin D1^-/p27^- patients against cyclin D1^-/p27⁺.

View larger version (18K): [in this window] [in a new window]   Fig 4. Overall survival of the patients classified on the basis of p27 and cyclin D1 expression (cyclin D1-/p27+, 59 patients; cyclin D1+/p27+–cyclin D1-/p27-, 51 patients; D1+/p27-, 22 patients).

	DISCUSSION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
In this study, we investigated by immunohistochemistry the clinical relevance of the expression of the members of the CIP/KIP family of cell-cycle inhibitors, p21 and p27, in a homogeneous series of 132 laryngeal carcinoma patients.

The clinical relevance of p21 expression in human tumors is still controversial: low levels of p21 expression have been found to correlate with a worse prognosis in lung,¹⁷ breast,¹⁸ prostate,²⁰ gastric,²² bladder,³⁹ uterine cervix,⁴⁰ and, at a borderline significant level, endometrial³³ carcinomas, but not in colorectal,³² pancreatic,⁴¹ and hepatocellular⁴² adenocarcinoma. With regard to head and neck tumors, high levels of p21 expression have been found to correlate with a worse clinical course in a small series of 42 tumors of the head and neck region,²⁶ but no association with prognosis was found in studies of the pyriform sinus²⁴ and larynx carcinomas.²⁵ We did not find any prognostic relevance of p21 expression in terms of disease-free and overall survival, despite the fact that low p21 levels were associated with the presence of lymph node metastases (P = .028), which is considered to be a predictor of poor prognosis in laryngeal cancer. However, in our series, this parameter did not show prognostic relevance on disease-free and overall survival. The reason for this may be that our therapeutic approach included postsurgical adjuvant locoregional radiotherapy for the subgroup of patients with lymph node metastases. Furthermore, the association between low p21 levels and advanced clinical stages, which turned out to have a significantly worse outcome, was above the statistical significance threshold (P = .089). Finally, the disease-free and overall survival curves of the patients with low levels of p21 lay just above those of patients with high levels, although the pattern was very similar (P = .6157 and P = .6797).

We found a different immunostaining pattern between p21 and p27 expression. In normal epithelium, p21 was expressed within the intermediate layers, whereas the basal and superficial layers were usually unreactive. Similar findings have been reported in colonic, small intestine, gastric, and esophageal mucosae,⁴³ where p21 expression is limited to the postreplicative compartment. Moreover, in accordance with recently reported findings in laryngeal cancer,⁴⁴ we found that p21 expression was significantly reduced in poorly differentiated carcinomas (P = .001 and P = .001 for trend) and more pronounced in the keratinizing cells of neoplastic nodules, a pattern closely resembling that observed in normal mucosa. Taken together, these data support the hypothesis that, as has already been found in in vitro studies,⁴⁵ p21 may also be involved in the process of cell differentiation in vivo. On the contrary, p27 was expressed throughout the laryngeal epithelium, including the basal proliferative compartment. This may seem to be somewhat surprising, because p27 is known to be a negative regulator of the cell cycle,¹⁰ but it is nevertheless in keeping with the fact that p27 expression has been detected in the basal layers of colonic,³¹ prostatic,⁴⁶ and bronchial¹⁵ mucosae, in exponentially growing human cancer cell lines,⁴⁷ and in a large proportion of human tumors,^{17,19,21,23,31,46} thus suggesting that replicative cells may tolerate high levels of p27 expression.

In a subset of cases (and consequently at a lower power level), we investigated the relationship between p21 expression and p53 gene mutations. We did not find any statistically significant association between p21 protein expression and p53 gene status (P = .648), although high p21 levels were more frequent in the cases with p53 mutations (61.5%) than in those bearing the wild-type p53 gene (48.2%). Our data are in accordance with those obtained from in vitro studies,^8,9 as well as from studies of different tumor types,^{22,32,33,44,48,49} and further confirm that the induction of p21 protein expression in laryngeal cancer may be regulated by p53-independent pathways.

p27 is a potential prognostic factor because its downregulation has recently been associated with a poor prognosis in a number of different malignancies, including non–small-cell lung,¹⁵ colon,¹⁶ breast,¹⁹ prostate,²¹ gastric,²³ and ovarian³⁴ carcinomas, malignant melanomas,⁵⁰ and non-Hodgkin's lymphomas.⁵¹ In this regard, it has to be underscored that there is no complete agreement in the literature concerning its division into classes, but most authors have used a cutoff point of less than 50% of immunoreactive neoplastic cells in defining low expressors. In our study, the reliability of this scoring method was further confirmed by the observation that, as in the cases of oral mucosa,⁵² the median value of p27 expression in morphologically normal mucosa was 49%. Using a two-class subdivision, we found that low p27 expression was significantly associated with the unfavorable clinicopathologic parameters of tumor extension and advanced clinical stage, as well as with tumor recurrence and reduced overall survival. Furthermore, in accordance with previous studies,^15,16,19 we also considered a classification that divided p27 low expressors into two subclasses (< 5% and 5% to 50%). In this case, the worst prognosis was found in the small subgroup of patients with very low p27 expression (< 5%), thus suggesting that a three-class subdivision may be a better model for assessing the clinical relevance of p27 expression in human tumors, although this has to be confirmed in larger series of patients.

p27 is capable of blocking progression from the G1 to the S phase of the cell cycle by binding cyclin E–cdk2 and cyclin A–cdk2.¹⁰ D-type cyclins and cdk4 competitively bind and downregulate the activity of p27 and may thereby act on a pathway that reverses cdk2 inhibition and enables G1 progression.⁵³ We found a statistically borderline inverse correlation between p27 and cyclin D1 expression (P = .092). In this regard, it has been reported that in mantle-cell lymphomas that constitutively express cyclin D1 as a result of a t(11;14) translocation, p27 protein is immunohistochemically undetectable because of binding by cyclin D1.⁵⁴ This suggests that the absence of p27 protein expression in vivo might be due to sequestering by cyclin D1 and that the balance between these two opposing regulators of the cell cycle may be the determining factor in cell proliferation. We analyzed the clinical behavior of patients classified on the basis of p27 and cyclin D1 expression and found that the group of patients with low levels of cyclin D1 and high levels of p27 (a phenotype recapitulating that observed in normal mucosa) had the best disease-free and overall survival probability; we also found that the group of patients with the opposite pattern (high levels of cyclin D1 expression and low levels of p27 expression) had the poorest disease-free and overall survival probability. These findings are further supported by the presence of a trend in both disease-free and overall survival (the patients with low levels of cyclin D1 and p27, as well as those with high levels of cyclin D1 and p27, had an intermediate survival probability) and lead to the speculation that multiple alterations in the genes involved in cell-cycle control are associated with greater tumor aggressiveness.

In conclusion, our data suggest that the immunohistochemical evaluation of p27 expression in laryngeal carcinomas may be a further useful prognostic marker for selecting subgroups of patients with a low prognostic score who can be treated with more aggressive surgical approaches, adjuvant radiotherapy, and chemotherapy, as well as for identifying those who require closer follow-up.

	ACKNOWLEDGMENTS

 
Supported by grants from the Associazione Italiana Ricerca sul Cancro (to A.N.) and from IRCCS Ospedale Maggiore of Milan (to R.B.).

The authors are indebted to Barbara Digiuni and Carmelo Arizzi for their excellent technical assistance.

	REFERENCES

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
1. Sherr CJ: Cancer cell cycles. Science274:1672-1677, 1996[Abstract/Free Full Text]

2. Harper JW: Cyclin dependent kinase inhibitors. Cancer Surv29:91-107, 1997[Medline]

3. Harper JW, Adami G, Wei N, et al: The p21kd Cdk interacting protein Cip1 is a potent inhibitor of G₁ cyclin-dependent kinases. Cell75:805-816, 1993[Medline]

4. Polyak K, Lee M-H, Erdjument-Bromage A, et al: Cloning of p27^KIP1, a cyclin-dependent kinase inhibitor and a potential mediator of extracellular antimitogenic signals. Cell78:59-66, 1994[Medline]

5. Matsuoka S, Edwards M, Bai C, et al: p57^Kip2, a structurally distinct member of the Cip1 cdk inhibitor family, is a candidate tumor suppressor gene. Genes Dev9:650-662, 1995[Abstract/Free Full Text]

6. El-Deiry WS, Tokino T, Velculescu VE, et al: WAF1, a potential mediator of p53 tumor suppression. Cell75:817-825, 1993[Medline]

7. El-Deiry WS, Harper JW, O'Connor PM, et al: WAF1/CIP1 is induced in p53-mediated G₁ arrest and apoptosis. Cancer Res54:1169-1174, 1994[Abstract/Free Full Text]

8. Michieli P, Chedid M, Lin D, et al: Induction of WAF1/CIP1 by a p53-independent pathway. Cancer Res54:3391-3395, 1994[Abstract/Free Full Text]

9. Zeng Y-X, El-Deiry WS: Regulation of p21^WAF1/CIP1 expression by p53-independent pathways. Oncogene12:1557-1664, 1996[Medline]

10. Polyak K, Kato J-Y, Solomon MJ, et al: p27^Kip1, a cyclin-Cdk inhibitor, links transforming growth factor-beta and contact inhibition to cell cycle arrest. Genes Dev8:9-22, 1994[Abstract/Free Full Text]

11. Shiohara M, El-Deiry WS, Wada M, et al: Absence of WAF1 mutations in a variety of human malignancies. Blood84:3781-3784, 1994[Abstract/Free Full Text]

12. Ponce-Castaneda MV, Lee MH, Latres E, et al: p27^kip1: Chromosomal mapping to 12p12-12p13.1 and absence of mutations in human tumors. Cancer Res55:1211-1214, 1995[Abstract/Free Full Text]

13. Li X-S, Rishi AK, Shao Z-M, et al: Posttranscriptional regulation of p21^WAF1/CIP1 expression in human breast carcinoma cells. Cancer Res56:5055-5062, 1996[Abstract/Free Full Text]

14. Hengst L, Reed SI: Translational control of p27^kip1 accumulation during the cell cycle. Science272:1861-1864, 1996

15. Esposito V, Baldi A, De Luca A, et al: Prognostic role of the cyclin-dependent kinase inhibitor p27 in non-small cell lung cancer. Cancer Res57:3381-3385, 1997[Abstract/Free Full Text]

16. Loda M, Cukor B, Tam SW, et al: Increased proteasome-dependent degradation of the cyclin-dependent kinase inhibitor p27 in aggressive colorectal carcinomas. Nat Med3:231-234, 1997[Medline]

17. Komiya T, Hosono Y, Hirashima N, et al: p21 expression as a predictor for favorable prognosis in squamous cell carcinoma of the lung. Clin Cancer Res3:1831-1835, 1997[Abstract]

18. Barbareschi M, Caffo O, Doglioni C, et al: p21^WAF1 immunohistochemical expression in breast carcinoma: Correlations with clinicopathological data, oestrogen receptor status, MIB1 expression, p53 gene and protein alterations and relapse-free survival. Br J Cancer74:208-215, 1996[Medline]

19. Porter PL, Malone KE, Heagerty PJ, et al: Expression of the cell-cycle regulators p27^Kip1 and cyclin E, alone and in combination, correlate with survival in young breast cancer patients. Nat Med3:222-225, 1997[Medline]

20. Matsushima H, Sasaki T, Goto T, et al: Immunohistochemical study of p21^WAF1 and p53 proteins in prostatic cancer and their prognostic significance. Hum Pathol29:778-783, 1998[Medline]

21. Cote RJ, Shi Y, Groshen S, et al. Association of p27^kip1 levels with recurrence and survival in patients with stage C prostate carcinoma. J Natl Cancer Inst 90:916-920, 1998

22. Gomyo Y, Ikeda M, Osaki M, et al: Expression of p21 (waf1/cip1/sdi1), but not p53 protein, is a factor in the survival of patients with advanced gastric carcinoma. Cancer79:2067-2072, 1997[Medline]

23. Mori M, Mimori K, Shiraishi T, et al: p27 expression and gastric carcinoma. Nat Med3:593, 1997 (letter) [Medline]

24. Nibu KI, Yanagisawa A, Nakamizo M, et al: Clinical role of p53 and p21^WAF1/CIP1 in squamous cell carcinoma of the pyriform sinus. Acta Otolaryngol118:432-437, 1998[Medline]

25. Jin Y-T, Kayser S, Kemp BL, et al: The prognostic significance of the biomarkers p21^WAF1/CIP1, p53, and bcl-2 in laryngeal squamous cell carcinoma. Cancer82:2159-2165, 1998[Medline]

26. Erber R, Klein W, Andl T, et al: Aberrant p21 (CIP1/WAF1) protein accumulation in head-and-neck cancer. Int J Cancer74:383-389, 1997[Medline]

27. Pruneri G, Pignataro L, Carboni N, et al: MDM-2 oncoprotein overexpression in laryngeal squamous cell carcinoma: Association with wild-type p53 accumulation. Mod Pathol10:785-792, 1997[Medline]

28. Pignataro L, Pruneri G, Carboni N, et al: Clinical relevance of cyclin D1 protein overexpression in laryngeal squamous cell carcinoma. J Clin Oncol16:3069-3077, 1998[Abstract/Free Full Text]

29. Hermanek P, Sobin LH (eds): UICC: TNM Classification of Malignant Tumors. Berlin, Germany, Springer Verlag, 1987

30. Shanmugaratnam K, Sobin IN, Pathologists in 8 Countries: Histological Typing of Tumours of the Upper Respiratory Tract and Ear. Berlin, Springer Verlag, 1987

31. Ciaparrone M, Yamamoto H, Yao Y, et al: Localization and expression of p27^KIP1 in multistage colorectal carcinogenesis. Cancer Res58:114-122, 1998[Abstract/Free Full Text]

32. Slebos RJC, Baas IO, Clement M, et al: Clinical and pathological associations with p53 tumor-suppressor gene mutations and expression of p21^WAF1/Cip1 in colorectal carcinoma. Br J Cancer74:165-171, 1996[Medline]

33. Ito K, Sasano H, Matsunaga G, et al: Correlations between p21 expression and clinicopathologic findings, p53 gene and protein alterations, and survival in patients with endometrial carcinoma. J Pathol183:318-324, 1997[Medline]

34. Newcombe EW, Sosnow M, Demopoulos RI, et al: Expression of the cell cycle inhibitor p27^KIP1 is a new prognostic marker associated with survival in epithelial ovarian tumors. Am J Pathol154:119-125, 1999[Abstract/Free Full Text]

35. Armitage P, Berry G: Statistical Methods in Medical Research. Oxford, United Kingdom, Blackwell Scientific Publications, 1994

36. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc53:457-481, 1958

37. Mantel N: Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep50:163-170, 1966[Medline]

38. Cox DR: Regression models and life-tables. J R Stat Soc B34:187-220, 1972

39. Stein JP, Ginsberg DA, Grossfeld GD, et al: Effect of p21^WAF1/CIP1 expression on tumor progression in bladder cancer. J Natl Cancer Inst90:1072-1079, 1998[Abstract/Free Full Text]

40. Lu X, Toki T, Konishi I, et al: Expression of p21^WAF1/CIP1 in adenocarcinoma of the uterine cervix: A possible immunohistochemical marker of a favorable prognosis. Cancer82:2409-2417, 1998[Medline]

41. Coppola D, Lu L, Fruehauf JP, et al: Analysis of p53, p21^WAF1, and TGF-beta1 in human ductal adenocarcinoma of the pancreas: TGF-beta1 protein expression predicts longer survival. Am J Clin Pathol110:16-23, 1998[Medline]

42. Naka T, Toyota N, Kaneko T, et al: Protein expression of p53, p21^WAF1, and Rb as prognostic indicators in patients with surgically treated hepatocellular carcinoma. Anticancer Res18:555-564, 1998[Medline]

43. El-Deiry WS, Tokino T, Waldman T, et al: Topological control of p21^WAF1/CIP1 expression in normal and neoplastic tissues. Cancer Res55:2910-2919, 1995[Abstract/Free Full Text]

44. Nadal A, Jares P, Cazorla M, et al: p21^WAF1/Cip1 expression is associated with cell differentiation but not with p53 mutations in squamous cell carcinomas of the larynx. J Pathol183:156-163, 1997[Medline]

45. Cox LS: Multiple pathways control cell growth and transformation: Overlapping and independent activities of p53 and p21^{Cip1/WAF1/Sdi1}. J Pathol183:134-140, 1998

46. Tsihlias J, Kapusta LR, DeBoer G, et al: Loss of cyclin-dependent kinase inhibitor p27^Kip1 is a novel prognostic factor in localized human prostate adenocarcinoma. Cancer Res58:542-548, 1998[Abstract/Free Full Text]

47. Fredersdorf S, Burns J, Milne AM, et al: High level expression of p27^Kip1 and cyclin D1 in some human breast cancer cells: Inverse correlation between the expression of p27^Kip1 and degree of malignancy in human breast and colorectal cancers. Proc Natl Acad Sci U S A94:6380-6385, 1997[Abstract/Free Full Text]

48. Di Giuseppe JA, Redston MS, Yeo CJ, et al: p53-independent expression of the cyclin-dependent kinase inhibitor p21 in pancreatic carcinoma. Am J Pathol147:884-888, 1995[Abstract]

49. Marchetti A, Doglioni C, Barbareschi M, et al: p21 RNA and protein expression in non-small cell lung carcinomas: Evidence of p53-independent expression and association with tumoral differentiation. Oncogene12:1319-1324, 1996[Medline]

50. Florenes VA, Maelandsmo GM, Kerbel RS, et al: Protein expression of the cell-cycle inhibitor p27^kip1 in malignant melanoma: Inverse correlation with disease-free survival. Am J Pathol153:305-312, 1998[Abstract/Free Full Text]

51. Erlanson M, Portin C, Linderholm B, et al: Expression of cyclin E, and the cyclin-dependent kinase inhibitor p27 in malignant lymphomas: Prognostic implications. Blood92:770-777, 1998[Abstract/Free Full Text]

52. Jordan RCK, Bradley G, Slingerland J: Reduced levels of the cell-cycle inhibitor p27^kip1 in epithelial dysplasia and carcinoma of the oral cavity. Am J Pathol152:585-590, 1998[Abstract]

53. Toyoshima H, Hunter T: p27, a novel inhibitor of G1 cyclin-Cdk protein kinase activity, is related to p21. Cell78:67-74, 1994[Medline]

54. Quintanilla-Martinez L, Davies-Hill T, Sorbara L, et al: Loss of detectable p27 expression in mantle cell lymphomas (MCL) is a result of its sequestration by cyclin D1: Implications for pathogenesis. Blood 92:314(A), 1998 (abstr)

Submitted March 11, 1999; accepted June 16, 1999.

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

This article has been cited by other articles:

				P. Sun, Y. Qiu, Z. Zhang, J. Wan, T. Wang, X. Jin, Q. Lan, N. Rothman, and Z.-l. Xia Association of Genetic Polymorphisms, mRNA Expression of p53 and p21 with Chronic Benzene Poisoning in a Chinese Occupational Population Cancer Epidemiol. Biomarkers Prev., June 1, 2009; 18(6): 1821 - 1828. [Abstract] [Full Text] [PDF]

				G. Deep, N. H. Oberlies, D. J. Kroll, and R. Agarwal Isosilybin B and isosilybin A inhibit growth, induce G1 arrest and cause apoptosis in human prostate cancer LNCaP and 22Rv1 cells Carcinogenesis, July 1, 2007; 28(7): 1533 - 1542. [Abstract] [Full Text] [PDF]

				M. D. Begnami, E. J. Rushing, R. Evangelista, M. Santi, and M. Quezado Evaluation of RB Gene and Cyclin-Dependent Kinase Inhibitors P21 and P27 in Pleomorphic Xantoastrocytoma International Journal of Surgical Pathology, April 1, 2006; 14(2): 113 - 118. [Abstract] [PDF]

				A. Celetti, D. Testa, S. Staibano, F. Merolla, V. Guarino, M. D. Castellone, R. Iovine, G. Mansueto, P. Somma, G. De Rosa, et al. Overexpression of the Cytokine Osteopontin Identifies Aggressive Laryngeal Squamous Cell Carcinomas and Enhances Carcinoma Cell Proliferation and Invasiveness Clin. Cancer Res., November 15, 2005; 11(22): 8019 - 8027. [Abstract] [Full Text] [PDF]

				G. Li, Z. Liu, E. M. Sturgis, Q. Shi, R. M. Chamberlain, M. R. Spitz, and Q. Wei Genetic polymorphisms of p21 are associated with risk of squamous cell carcinoma of the head and neck Carcinogenesis, September 1, 2005; 26(9): 1596 - 1602. [Abstract] [Full Text] [PDF]

				Y. Dong, L. Sui, Y. Watanabe, F. Yamaguchi, N. Hatano, and M. Tokuda Prognostic Significance of Jab1 Expression in Laryngeal Squamous Cell Carcinomas Clin. Cancer Res., January 1, 2005; 11(1): 259 - 266. [Abstract] [Full Text] [PDF]

				M. Rodriguez-Pinilla, J. L. Rodriguez-Peralto, R. Hitt, J. J. Sanchez, C. Ballestin, A. Diez, L. Sanchez-Verde, F. Alameda, and M. Sanchez-Cespedes Cyclin A as a Predictive Factor for Chemotherapy Response in Advanced Head and Neck Cancer Clin. Cancer Res., December 15, 2004; 10(24): 8486 - 8492. [Abstract] [Full Text] [PDF]

				X. Xie, O. P. F. Clausen, and M. Boysen Prognostic Significance of p21WAF1/CIP1 Expression in Tongue Squamous Cell Carcinomas Arch Otolaryngol Head Neck Surg, August 1, 2002; 128(8): 897 - 902. [Abstract] [Full Text] [PDF]

				A. M. Oliveira, A. G. Nascimento, S. H. Okuno, and R. V. Lloyd p27kip1 Protein Expression Correlates With Survival in Myxoid and Round-Cell Liposarcoma J. Clin. Oncol., August 15, 2000; 18(15): 2888 - 2893. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Pruneri, G. Articles by Neri, A. Search for Related Content PubMed PubMed Citation Articles by Pruneri, G. Articles by Neri, A. Social Bookmarking                            What's this?