1

Introduction

Sinonasal inverted papilloma (IP) is a benign disease, but its potential for malignant transformation is 6% to 15% . Although studies have examined the roles of human papilloma virus infection , apoptosis , and cell adhesion molecules in sinonasal IP, the mechanism of the malignant transformation of benign IP into squamous cell carcinoma (SCC) has not been fully elucidated.

Current research on the mechanisms of carcinogenesis is focused on 2 issues: the imbalance between cell proliferation and apoptosis, and inflammation-mediated oncogenesis. Proapoptotic factors such as p53, p21, and p16 and antiapoptotic factors such as p63, and the bcl-2 family have been investigated in several neoplasms including sinonasal IP . In addition, chronic inflammation is a well-established risk factor for several types of cancer, particularly in the gastrointestinal tract , and the causal relationship between inflammation and cancer is known.

The most extensively investigated bridge molecules between chronic inflammation and carcinogenesis are cyclooxygenase-2 (COX-2) and nuclear factor κ B (NF- κ B). Cyclooxygenase-2 is not detectable in most normal tissues, is induced by mitogenic and inflammatory stimuli, and promotes tumorigenesis by inducing neoangiogenesis, inhibiting apoptosis, and altering cell adhesion molecules . It is overexpressed in premalignant and malignant tissues , including those in the head and neck .

Nuclear factor κ B, a ubiquitous nuclear transcription factor, plays a major role in inflammation. It most commonly exists as a p65/p50 heterodimer, which is retained in the cytoplasm by the inhibitory subunit I κ B α . The stimulation of cells with various inflammatory and oxidative stresses activates I κ B α kinase, leading to the sequential phosphorylation and degradation of I κ B α and release of the p50/p65 heterodimer. Nuclear factor κ B is then translocated to the nucleus, where it binds to multiple DNA sequences to initiate the transcription of a wide variety of genes including inflammatory cytokines, angiogenesis factors, cell adhesion molecules, and antiapoptotic factors . Recently, a series of studies showed that NF- κ B activation is associated with the initiation and progression of several human cancers, indicating that it may be a molecular link between chronic inflammation and cancer development .

It has been suggested that sinonasal IP arises from a background of sinonasal mucosal inflammation . A system for grading sinonasal IP, which ranges from inflammatory polyps with squamous metaplasia to SCC, according to the extent of inflammatory infiltration and cellular atypia has been proposed. A histologic examination identified a significant inflammatory cell population in sinonasal IP, and early-grade IP lesions had more inflammatory cells than did advanced lesions such as IP with dysplasia or SCC. Therefore, inflammation was suggested to be the initiating step in the histologic transformation of sinonasal IP. However, no study has examined the role of chronic inflammation in the cellular transformation of benign sinonasal IP into advanced lesions.

This study evaluated the roles of the bridge molecules COX-2 and NF- κ B along with apoptosis-related proteins in the transformation of early-grade sinonasal IP into advanced-grade IP.

2

Materials and methods

2.2

Methods

2.2.1

Histopathologic grading of sinonasal IPs

Original H&E-stained slides were reviewed and graded by one pathologist (J.Y. Kim) using the scheme reported by Roh et al ( Fig. 1 ). Based on the grading system, the most representative portion of the specimen was marked on the slide. When different grades were identified on the same slide, a representative portion of each grade was marked. For example, only dysplastic foci were selected for grade III lesions, and only areas of invasive SCC were selected for grade IV lesions. Grade II lesions were subdivided into 2 groups: lesions from IP without dysplasia or SCC (group A) and lesions from IP with dysplasia or SCC (group B).

Representative examples of COX-2 staining in sinonasal IP. Each figure shows negative cytoplasmic staining in grade I (A), weak cytoplasmic staining (1 +) in grade II (B), moderate cytoplasmic staining (2 +) in grade III (C), and strong cytoplasmic staining (3 +) in grade IV (D). Normal nasal respiratory mucosa shows mild cytoplasmic staining in respiratory epithelial cells (arrow head) (original magnification × 100).

2.2.2

Tissue microarray

The marked H&E-stained slide was placed on the surface of the donor paraffin block, and the marked region was defined on the donor block. Using 2-mm punch needles, tissue cores were punched from the defined regions on the donor block and transferred to ready-made holes in a recipient paraffin block, at defined x – y positions. The resulting tissue microarray (TMA) was placed upside down, covered with melted paraffin, and warmed in an oven for 5 minutes at 50°C. The paraffin of the tissue cores and recipient block started to homogenize, resulting in the final array block, and histologic sections were prepared from the TMA. From the 118 specimens, 85 tissue cores were obtained. Of these, only 80 cores were evaluated because 5 cores were lost during the staining procedure. Respectively, 32, 37, 6, and 5 belonged to grades I, II, III, and IV lesions ( Table 1 ).

Table 1

Histopathologic grading of IP specimens

Grade	Histopathologic characteristics	n (%)
I	Ciliated respiratory epithelium with underlying squamous metaplasia, at least 12 cell layers thick, associated with early inversion of squamous metaplasia	32 (40.0)
II	Partially ciliated respiratory epithelium with luminal squamous metaplasia and increased prominence of inversion including the involvement of seromucinous glands	37 (46.2)
III	Near-complete absence of respiratory epithelium, metaplasia replaced by stratified squamous epithelium, and varying degrees of dysplasia	6 (7.5)
IV	Invasive SCC in the presence of inverted squamous metaplastic and dysplastic change; features of grades II and II will be present when grade IV is diagnosed	5 (6.3)
Total		80 (100)

2.2.3

Immunohistochemistry

Tissue microarray sections of 4- μ m thickness were mounted on slide glasses and air dried overnight. The sections were deparaffinized in xylene twice for 5 minutes each, rehydrated in graded ethanol baths (100%, 100%, 95%, 80%, and 70%) for 5 minutes each, and washed thoroughly in distilled water (DW) for 10 minutes. After heat-mediated antigen retrieval in a micro-oven, using TE (Tris-EDTA) buffer (10 mmol/L Tris-HCl, 1 mmol/L EDTA) for COX-2 and 0.01 mol/L citrate buffer for p53, NF- κ B, bax, and bcl-2, the sections were allowed to cool for 5 minutes at room temperature (RT), followed by washing in DW for 30 seconds and in phosphate-buffered saline (PBS) for 10 minutes. The slide glass was marked with a PAP pen (Invitrogen, Grand Island, NY) on either side of the section, and endogenous peroxidase activity was blocked with 3% H ₂ O ₂ in DW for 15 minutes. The sections were again washed in DW for 30 seconds and in PBS for 10 minutes, and then incubated with a primary antibody for 1 hour at RT. The primary antibodies included mouse antihuman COX-2 (clone CX-294, dilution 1:50; Dako, Copenhagen, Denmark), mouse antihuman p53 (clone DO-7, dilution 1:100; Dako), and mouse antihuman bcl-2 monoclonal antibodies (clone 124, dilution 1:100; Dako) as well as rabbit antihuman bax (dilution 1:200; Dako) and rabbit antihuman NF- κ B polyclonal antibodies (dilution 1:50; Santa Cruz Biotechnology, Santa Cruz, CA). After a wash in PBS for 10 minutes, the sections were incubated for 30 minutes with a horseradish peroxidase–labeled polymer (Dako EnVision + System) and washed in PBS for 10 minutes. They were incubated in EnVision 3,3-diaminobenzidine chromogen solution (Dako) at RT for 2 minutes, resulting in a brown precipitate at the antigen site, and then placed in running tap water for 5 minutes. The sections were counterstained with Mayer hematoxylin for 1 minute, dehydrated in a graded ethanol series (95%, 100%, 100%, 100%), and cleared in xylene. Cover slides were placed on the sections using a resinous mountant, and the slides were left to dry at RT.

2.2.4

Positive criteria for immunohistochemical staining

For p53 and NF- κ B, only clear nuclear staining was counted as positive. The percentage of positive cells was counted for p53 and graded on a 4-point scale: less than 10% (0), 10% or more and less than 50% (1 +), 50% or more and less than 90% (2 +), and 90% or more (3 +). For NF- κ B, tissue sections showing 10% or more cells with nuclear immunostaining were scored as positive. For COX-2, bax, and bcl-2, the presence of cytoplasmic staining was considered positive, and the immunopositivity was graded as none (0), weak (1 +), moderate (2 +), or strong (3 +) according to the staining intensity.

2.2.5

Statistical analysis

The correlations between histologic grade and the expression of each protein were analyzed using Spearman correlation analysis. The Wilcoxon rank sum test was used to compare protein expression in each grade. Statistical analyses were performed using SPSS 17.0 (SPSS Inc, Chicago, IL), and the results were considered statistically significant at P < .05.

2

Materials and methods

2.2

Methods

2.2.1

Histopathologic grading of sinonasal IPs

Original H&E-stained slides were reviewed and graded by one pathologist (J.Y. Kim) using the scheme reported by Roh et al ( Fig. 1 ). Based on the grading system, the most representative portion of the specimen was marked on the slide. When different grades were identified on the same slide, a representative portion of each grade was marked. For example, only dysplastic foci were selected for grade III lesions, and only areas of invasive SCC were selected for grade IV lesions. Grade II lesions were subdivided into 2 groups: lesions from IP without dysplasia or SCC (group A) and lesions from IP with dysplasia or SCC (group B).

Representative examples of COX-2 staining in sinonasal IP. Each figure shows negative cytoplasmic staining in grade I (A), weak cytoplasmic staining (1 +) in grade II (B), moderate cytoplasmic staining (2 +) in grade III (C), and strong cytoplasmic staining (3 +) in grade IV (D). Normal nasal respiratory mucosa shows mild cytoplasmic staining in respiratory epithelial cells (arrow head) (original magnification × 100).

2.2.2

Tissue microarray

The marked H&E-stained slide was placed on the surface of the donor paraffin block, and the marked region was defined on the donor block. Using 2-mm punch needles, tissue cores were punched from the defined regions on the donor block and transferred to ready-made holes in a recipient paraffin block, at defined x – y positions. The resulting tissue microarray (TMA) was placed upside down, covered with melted paraffin, and warmed in an oven for 5 minutes at 50°C. The paraffin of the tissue cores and recipient block started to homogenize, resulting in the final array block, and histologic sections were prepared from the TMA. From the 118 specimens, 85 tissue cores were obtained. Of these, only 80 cores were evaluated because 5 cores were lost during the staining procedure. Respectively, 32, 37, 6, and 5 belonged to grades I, II, III, and IV lesions ( Table 1 ).

Table 1

Histopathologic grading of IP specimens

Grade	Histopathologic characteristics	n (%)
I	Ciliated respiratory epithelium with underlying squamous metaplasia, at least 12 cell layers thick, associated with early inversion of squamous metaplasia	32 (40.0)
II	Partially ciliated respiratory epithelium with luminal squamous metaplasia and increased prominence of inversion including the involvement of seromucinous glands	37 (46.2)
III	Near-complete absence of respiratory epithelium, metaplasia replaced by stratified squamous epithelium, and varying degrees of dysplasia	6 (7.5)
IV	Invasive SCC in the presence of inverted squamous metaplastic and dysplastic change; features of grades II and II will be present when grade IV is diagnosed	5 (6.3)
Total		80 (100)

2.2.3

Immunohistochemistry

Tissue microarray sections of 4- μ m thickness were mounted on slide glasses and air dried overnight. The sections were deparaffinized in xylene twice for 5 minutes each, rehydrated in graded ethanol baths (100%, 100%, 95%, 80%, and 70%) for 5 minutes each, and washed thoroughly in distilled water (DW) for 10 minutes. After heat-mediated antigen retrieval in a micro-oven, using TE (Tris-EDTA) buffer (10 mmol/L Tris-HCl, 1 mmol/L EDTA) for COX-2 and 0.01 mol/L citrate buffer for p53, NF- κ B, bax, and bcl-2, the sections were allowed to cool for 5 minutes at room temperature (RT), followed by washing in DW for 30 seconds and in phosphate-buffered saline (PBS) for 10 minutes. The slide glass was marked with a PAP pen (Invitrogen, Grand Island, NY) on either side of the section, and endogenous peroxidase activity was blocked with 3% H ₂ O ₂ in DW for 15 minutes. The sections were again washed in DW for 30 seconds and in PBS for 10 minutes, and then incubated with a primary antibody for 1 hour at RT. The primary antibodies included mouse antihuman COX-2 (clone CX-294, dilution 1:50; Dako, Copenhagen, Denmark), mouse antihuman p53 (clone DO-7, dilution 1:100; Dako), and mouse antihuman bcl-2 monoclonal antibodies (clone 124, dilution 1:100; Dako) as well as rabbit antihuman bax (dilution 1:200; Dako) and rabbit antihuman NF- κ B polyclonal antibodies (dilution 1:50; Santa Cruz Biotechnology, Santa Cruz, CA). After a wash in PBS for 10 minutes, the sections were incubated for 30 minutes with a horseradish peroxidase–labeled polymer (Dako EnVision + System) and washed in PBS for 10 minutes. They were incubated in EnVision 3,3-diaminobenzidine chromogen solution (Dako) at RT for 2 minutes, resulting in a brown precipitate at the antigen site, and then placed in running tap water for 5 minutes. The sections were counterstained with Mayer hematoxylin for 1 minute, dehydrated in a graded ethanol series (95%, 100%, 100%, 100%), and cleared in xylene. Cover slides were placed on the sections using a resinous mountant, and the slides were left to dry at RT.

2.2.4

Positive criteria for immunohistochemical staining

For p53 and NF- κ B, only clear nuclear staining was counted as positive. The percentage of positive cells was counted for p53 and graded on a 4-point scale: less than 10% (0), 10% or more and less than 50% (1 +), 50% or more and less than 90% (2 +), and 90% or more (3 +). For NF- κ B, tissue sections showing 10% or more cells with nuclear immunostaining were scored as positive. For COX-2, bax, and bcl-2, the presence of cytoplasmic staining was considered positive, and the immunopositivity was graded as none (0), weak (1 +), moderate (2 +), or strong (3 +) according to the staining intensity.

2.2.5

Statistical analysis

The correlations between histologic grade and the expression of each protein were analyzed using Spearman correlation analysis. The Wilcoxon rank sum test was used to compare protein expression in each grade. Statistical analyses were performed using SPSS 17.0 (SPSS Inc, Chicago, IL), and the results were considered statistically significant at P < .05.