Basal/myoepithelial cells in chronic sinusitis, respiratory epithelial adenomatoid hamartoma, inverted papilloma, and intestinal-type and nonintestinal-type sinonasal adenocarcinoma: an immunohistochemical study.

* Context.--The pathogenesis of respiratory epithelial adenomatoid hamartoma (REAH) and inverted papilloma (IP) is poorly understood, especially compared with sinonasal adenocarcinoma (SNAC). One feature of malignant glandular lesions is loss of the basal/myoepithelial layer. The immunophenotype of the basal/myoepithelial layer has not been fully examined in benign glandular lesions of the sinonasal tract.

Objective.--To examine benign and malignant glandular lesions in the sinonasal tract for the immunophenotype of basal/myoepithelial cells, proliferation index, and cytokeratin and intestinal differentiation profiles.

Design.--Sinonasal adenocarcinoma (intestinal-type adenocarcinoma [ITAC] and nonintestinal type adenocarcinoma [non-ITAC]), REAH, IP, and chronic sinusitis (CS) were stained for cytokeratin (CK) 7, CK20, 34[beta]E12, CDX-2, p63, Ki-67, smooth muscle actin (SMA), S100 protein, and calponin.

Results.--Basal/myoepithelial cells in CS and REAH were positive for p63 and 34[beta]E12 but negative for SMA, S100 protein, and calponin. Proliferative activity was localized to the compartment containing p63-positive cells. Inverted papilloma demonstrated broad areas staining for p63 and 34[beta]E12, with intermediate proliferative activity in these areas. Sinonasal adenocarcinoma had the highest Ki-67 labeling index, and p63-positive SNACs had higher proliferation indices than p63-negative SNACs. REAH, IP, CS, and most SNACs expressed CK7. Only SNAC expressed CK20. Sixty percent of morphologic ITACs expressed CDX-2.

Conclusions.--Basal/myoepithelial cells in CS and REAH should be considered basal and not myoepithelial cells. In benign lesions, proliferative activity is limited to the compartments with p63 staining. In SNAC and IP, p63 expression correlates with proliferation index. REAH, IP, and CS share similar immunoprofiles (CK[7.sup.+], CK[20.sup.-], and CDX-[2.sup.-]), contrasting with SNAC (CK[7.sup.+], CK[20.sup.+/-], CDX-[2.sup.-/+]).

(Arch Pathol Lab Med. 2007;131:530-537)

Benign and malignant glandular lesions of the sinonasal tract are uncommon and can present a diagnostic dilemma to the pathologist, particularly on small biopsy samples. If excluding common salivary gland tumors, most of the remaining glandular lesions include respiratory epithelial adenomatoid hamartoma (REAH) and sinonasal adenocarcinomas (SNACs), of the intestinal (ITAC) and nonintestinal (non-ITAC) types. Although inverted papilloma (IP) has a unique appearance, it is often confused with REAH and therefore was included in this study.

Respiratory epithelial adenomatoid hamartoma is an uncommon lesion that usually presents as a nasal or, infrequently, as a sinus or nasopharyngeal mass. In the largest series to date (31 cases), REAH was defined as an excessive proliferation of glandular respiratory epithelium and therefore considered to be a hamartoma. (1) Morphologically, it is composed of respiratory-lined glands surrounded by a thickened and hyalinized basement membrane, often accompanied by an inflammatory background. Although there are characteristic histologic and immunophenotypic features that distinguish between REAH, IP, ITAC, and non-ITAC, (2) the differential diagnosis can be challenging on small or fragmented endoscopic biopsies.

In other organ systems, the preservation of basal/myoepithelial cells is a useful feature to distinguish between benign, in situ, and invasive tumors. Both p63 and 34[beta]E12 are antibodies that have been used to label basal/myoepithelial cells in a variety of tissue types, including breast, prostate, salivary gland, and lower respiratory tract epithelium. (3-5) The basal/myoepithelial cell also may be seen in normal sinonasal mucosa, but the exact lineage is not well established. While p63 is used as a marker for basal and myoepithelial cells, other markers are needed to determine myoepithelial differentiation. In addition, p63 is also implicated by some investigators as a stem cell marker, and a link between p63 expression and the proliferative capacity of epithelial cells has been postulated. (6)

We undertook this study to (1) confirm prior studies of differentially expressed cytokeratins in ITACs and non-ITACs, especially in comparison to the profiles seen in benign glandular lesions; (2) define the immunophenotypic characteristics of basal/myoepithelial cells found in sinonasal mucosa and examine benign and malignant lesions of sinonasal mucosa using antibodies targeted at basal and myoepithelial lineage; and (3) compare proliferation rates and proliferative cellular compartments among sinonasal lesions using a Ki-67 labeling index (KLI).

MATERIALS AND METHODS

Case Selection

This study was approved by the institutional review board of the University of Pittsburgh as an exempt protocol (IRB #0310111). Cases with the diagnosis of REAH, IP, SNAC (to include 5 with the designation of ITAC and 5 with the designation of non-ITAC adenocarcinoma), and chronic sinusitis (CS) were identified for immunohistochemistry and were retrieved from the pathology archives at the University of Pittsburgh Medical Center from 1983 to 2005. Salivary gland-type adenocarcinomas were excluded from the sinonasal adenocarcinoma category. For IP and CS, 10 consecutive cases from 2004 to 2005 were selected. Blocks that were extensively decalcified were excluded. The hematoxylin-eosin-stained slides were reviewed, and the diagnosis was confirmed by one of the study pathologists (J.L.H. or E.L.B.) prior to inclusion.

Immunohistochemistry

The immunohistochemistry stains were performed on additional paraffin-embedded tissue sections. The antibodies used and the protocols for staining are listed in Table 1. All stains were examined by light microscopy, as were positive and negative controls. For all antibodies tested, positive and negative controls were appropriate. Qualitative analysis (positive vs negative or semiquantified where shown) was used for all antibodies except Ki-67, which was quantified. The KLI was calculated by dividing the number of staining nuclei for Ki-67 in the area of lesion with the highest density of cells staining positively, dividing by the total number of cells, and multiplying by 100 to obtain the percentage of positive cells (minimum of 1000 cells counted).

Statistical Analysis

All statistical analysis was performed using Stata Statistical Software, Release 8.0 (StataCorp LP, College Station, Tex). Age across the four groups was compared using a one-way analysis of variance. Sex was compared using an Exact chi-square test. Due to the small sample size, a Fisher exact test was used to compare the proportions of lesions demonstrating a particular immunophenotype. A Fisher exact test also was performed to identify statistically significant differences between proportions of cases that demonstrated positive staining for SNAC and REAH, SNAC and IP, and SNAC and CS for each stain. A Mann-Whitney test was used to compare the KLI% as a continuous variable to identify statistically significant differences between SNAC and REAH, SNAC and IP, and SNAC and CS. All tests were conducted as two-tail tests, with P values less than or equal to .05 considered statistically significant.

RESULTS

Clinical Findings

Age and sex of patients in each diagnostic group are shown in Table 2. No statistically significant differences were noted for age or sex between the differential diagnoses.

Histopathologic and Immunohistochemical Findings

Immunohistochemical staining results are shown in Tables 3 and 4. Representative hematoxylin-eosin images of REAH, IP, and SNAC are shown in Figure 1, A through D, respectively, and demonstrate the classic morphologic features for these lesions. Immunohistochemical results given below are annotated by lesion type.

[FIGURE 1 OMITTED]

Within the REAH group, the surface and stromal epithelial components were positive for cytokeratin (CK) 7 (Figure 2, A) and negative for CK20 and CDX-2. A basal cell layer could be appreciated that was maintained around the deep glands and was highlighted uniformly by both p63 (Figure 2, B) and 34[beta]E12 (Figure 2, C). Areas with squamous metaplasia, when present, were also uniformly positive for 34[beta]E12. Weaker cytoplasmic staining for 34[beta]E12 was seen in scattered surface ciliated cells as well. Ki-67 staining was localized to the basal/myoepithelial layer and demonstrated the lowest mean labeling index of any group of lesions at 0.3% (Figure 2, D). None of the epithelial or basal/myoepithelial cells stained with smooth muscle actin (SMA), calponin, or S100.

[FIGURE 2 OMITTED]

In IP, CK7 was strongly positive in areas of respiratory epithelium, whereas in squamous-appearing areas CK7 was only focally expressed (Figure 3, A and B). Both CDX-2 and CK20 were negative in the IP. p63 was positive in basal/myoepithelial cells in the respiratory epithelium areas and stained diffusely in other areas (Figure 3, C). Inverted papilloma showed a significantly higher number of lesions showing epithelial staining for p63 and 34[beta]E12 (Figure 3, D) than SNAC (Table 3). The mean KLI was 8.3% (Figure 3, E). Ki-67-positive cells and p63-positive cells seemed to localize to the same basilar regions of IP lesions. None stained with SMA, calponin, or S100.

[FIGURE 3 OMITTED]

Within the category of SNAC (ITAC and non-ITAC), the majority were positive for CK7 (90%; Figure 4, A) and half for CK20 (Figure 4, B). Overall, a significantly higher number of SNACs showed staining with CK20 than REAH, IP, or CS (Table 4). A total of 3 (60%) of 5 ITACs were CK20 positive (Figure 4, B), compared with 1 (20%) of 5 non-ITACs (Table 5). A total of 3 (60%) of 5 cases designated ITAC by histomorphology were CDX-2 positive (Figure 4, C). A total of 3 (60%) of 5 ITACs were both CK7 and CK20 positive. These 3 cases also were CDX-2 positive. A total of 3 (30%) of 10 SNACs showed staining with p63. One poorly differentiated (CK[7.sup.-], CK[20.sup.-], CDX-[2.sup.-]) non-ITAC showed diffuse strong staining (Figure 4, D), while one moderately differentiated mucinous ITAC (CK[7.sup.+], CK[20.sup.+], CDX-[2.sup.-]) and a low-grade papillary ITAC (CK[7.sup.+], CK[20.sup.-], CDX-[2.sup.-]) demonstrated scattered p63-positive cells within the tumor. A total of 50% of SNACs were 34[beta]E12 positive (both poorly differentiated non-ITACs and 3 ITACs; 1 ITAC had a rare positive cell; Table 5 and Figure 4, E), but none stained with SMA or calponin. A total of 3 SNACs (2 poorly differentiated non-ITACs and 1 well-differentiated ITAC) had focal staining for S100 protein (not shown). The mean KLI was highest among these lesions at 12.8% (Figure 4, F). Additionally, the KLI for p63-positive SNACs was higher than for p63-negative SNACs (27% vs 6%, respectively), but this did not reach statistical significance (P = .07).

[FIGURE 4 OMITTED]

The staining pattern seen in CS demonstrated that Schneiderian epithelium was positive for CK7 (Figure 5, A) but negative for CK20 and CDX-2. The basal cell component was highlighted by p63 (Figure 5, B) but not with SMA, calponin, or S100. All cases of CS demonstrated not only prominent basal and suprabasal staining with 34[beta]E12 (Figure 5, C), but also more widespread and more intense staining of the surface and luminal cells. Proliferative activity was limited but was higher than REAH, with a mean KLI of 4.7% (Figure 5, D), and like REAH was confined to the basal/myoepithelial cell layer.

[FIGURE 5 OMITTED]

COMMENT

[FIGURE 4 OMITTED]

Distinguishing between benign and malignant glandular lesions of the sinonasal tract can be challenging, particularly on small or fragmented biopsies. Histologic and immunohistochemical characteristics are, at times, helpful. However, reported studies often do not include the more recently recognized REAH or the immunoprofile of normal Schneiderian epithelium. Although the absence of a basal/myoepithelial compartment is a characteristic feature of malignant lesions of the sinonasal tract, the derivation of these cells and the implications in benign lesions is not well understood. This study was performed to assess the staining characteristics of the epithelial and basal cells in benign and malignant glandular lesions of the sinonasal tract, including REAH, IP, CS, and SNAC. Salivary gland-type lesions were excluded.

Many prior studies have used p63 and 34[beta]E12 as markers for basal and myoepithelial cells in different organ systems. (3-5) The typical myoepithelial cells seen in salivary glands will stain for other markers, such as calponin, SMA, and S100. (7,8) Although it is known that cultured human nasal epithelial cells are 34[beta]E12 negative, and isolated basal cells in cell culture are positive for 34[beta]E12, (9,10) it has not been well established whether these basal cells have any myoepithelial differentiation and whether they are seen in the benign glandular lesions of this region.

In this study, our results suggest that the basal cells do not have myoepithelial differentiation, as they are positive for p63 and uniformly negative for calponin, SMA, and S100 protein stains. In REAH and CS, only basal cells were positive for p63 and 34[beta]E12. In contrast, all of the IP stained more uniformly with p63 and 34[beta]E12. The ITAC and non-ITAC stained variably with p63 (3/10) and 34[beta]E12 (5/10).

p63 is not only used as a marker for differentiating possible cellular subtype, but it also has been linked to proliferative capacity. It is a member of the tumor suppressor p53 family of proteins that, in conjunction with a third member (p73), mediate cellular damage response. Due to their unique carboxyterminal extensions, p63 and p73 also play important primary roles in development, cellular differentiation, and cell proliferation. (11) Basilar expression of p63 in human oral mucosa has led some to postulate a connection between p63 expression and the proliferative capacity of the basal layer. (6) Our results support this association in that proliferation (by KI-67 staining) is present in the cellular compartment with p63-positive cells in REAH and CS. Conversely, in IP, ITACs, and non-ITACs, where more diffuse (nonbasilar) p63 staining was seen, these areas also demonstrated more diffuse Ki-67 staining.

We also examined the epithelium of the lesions for expression of commonly used cytokeratins (CK7 and CK20) and for CDX-2, a homeobox transcription factor important in intestinal differentiation. The epithelial component of REAH, CS, and IP showed strong expression of CK7 and was negative for CK20 and CDX-2. The cytokeratin profiles of ITACs and non-ITACs have been very well studied and generally show that between 67% and 100% of ITACs will express CK20, and many will express CK7, CK20, and CDX-2 in combination. (12-15) The percentage of ITACs showing CK20 expression in our current study (50%) generally is lower than in previously published articles, but a similar percentage of our cases expressed all 3 markers in combination (CK7, CK20, and CDX-2). Most ITACs will express CDX-2 compared with non-ITACs, and this is one marker proposed to distinguish ITACs from non-ITACs. (14,16) In this study, 60% of cases designated as ITACs were CDX-2 positive, whereas no non-ITACs were CDX-2 positive.

This study has several limitations. First, since several of the cases included in this series were consultation cases, long-term follow-up was not available, and therefore correlation between morphology, proliferative activity, and outcome was not possible. Second, because of the low number of each different type of ITAC and non-ITAC in our series, we were unable to draw any conclusions about the degree of p63 and Ki-67 staining and the cellular phenotype.

In summary, this study demonstrates that basal cells are present in the Schneiderian membrane and in REAH that stain with p63 and 34[beta]E12 but not with markers for myoepithelial differentiation. Invereted papilloma were diffusely positive for these markers. ITACs and non-ITACs have loss of the basal cell layer, but in some cases they show expression of p63 and 34[beta]E12 in the malignant epithelial component. Proliferative activity as seen by Ki-67 staining was seen in the compartment with p63-positive cells in all lesions. In our cases, CK20 and CDX-2 only were seen in most ITACs and not in REAH or IP. This staining pattern of p63, 34[beta]E12, and Ki-67 brings into question whether expression is related to the pathogenesis or is merely an association with the cellular phenotypes in these lesions. Further investigation with molecular studies will be essential to further our understanding of this pathogenesis.

The authors gratefully acknowledge Laura Cassidy, PhD, from the Department of Biostatistics at the Graduate School of Public Health of the University of Pittsburgh for the statistical analysis, and Kimberly Fuhrer, MT, for her technical expertise in performing the immunohistochemistry. We thank contributors of cases of respiratory epithelial adenomatoid hamartoma: Dmitriy Gutkin, PhD (Veterans Administration Hospital, Pittsburgh, Pa), Margaret Brandwein-Gensler, MD (Montefiore Medical Center, Albert Einstein School of Medicine, Bronx, NY), Douglas Richardson, MD (River Oaks Medical Laboratory, Jackson, Miss), Jagjit Singh, MD (UPMC St. Margaret's Hospital, Pittsburgh, Pa), and Ronald Berardi, MD (Latrobe Area Hospital, Latrobe, Pa).

References

(1.) Wenig BM, Heffner DK. Respiratory epithelial adenomatoid hamartomas of the sinonasal tract and nasopharynx: a clinicopathologic study of 31 cases. Ann Otol Rhinol Laryngol. 1995;104:639-645.

(2.) Barnes L, Brandwein M, Som PM. Diseases of the nasal cavity, paranasal sinuses, and nasopharynx. In: Barnes L, ed. Surgical Pathology of the Head and Neck. 2nd ed. New York, NY: Marcel Dekker, Inc; 2001:439-555.

(3.) Reis-Filho JS, Simpson PT, Martins A, Preto A, Gartner F, Schmitt FC. Distribution of p63, cytokeratins 5/6 and cytokeratin 14 in 51 normal and 400 neoplastic human tissue samples using TARP-4 multi-tumor tissue microarray. Virchows Arch. 2003;443:122-132.

(4.) O'Connell JT, Mutter GL, Cviko A, et al. Identification of a basal/reserve cell immunophenotype in benign and neoplastic endometrium: a study with the p53 homologue p63. Gynecol Oncol. 2001;80:30-36.

(5.) Barbareschi M, Pecciarini L, Cangi MG, et al. P63, a p53 homologue, is a selective nuclear marker of myoepithelial cells of the human breast. Am J Surg Pathol. 2001;25:1054-1060.

(6.) Lo Muzio L, Santarelli A, Caltabiano R, et al. P63 overexpression associates with poor prognosis in head and neck squamous cell carcinoma. Hum Pathol. 2005;36:187-194.

(7.) Ogawa Y. Immunocytochemistry of myoepithelial cells in the salivary glands. Prog Histochem Cytochem. 2003;38:343-426.

(8.) Furuse C, Sousa SO, Nunes FD, Magalhaes MH, Araujo VC. Myoepithelial cell markers in salivary gland neoplasms. Int J Surg Pathol. 2005;13:57-65.

(9.) McManus MS, Altman LC, Koenig JQ, et al. Human nasal epithelium: characterization and effects of in vitro exposure to sulfur dioxide. Exp Lung Res. 1989; 15:849-865.

(10.) Hicks W Jr, Hall L, Sigurdson L, et al. Isolation and characterization of basal cells from human upper respiratory epithelium. Exp Cell Res. 1997;237: 357-363.

(11.) Moll UM, Slade N. P63 and p73: roles in development and tumor formation. Mol Cancer Res. 2004;2:371-386.

(12.) Choi HR, Sturgis EM, Rashid A, et al. Sinonasal adenocarcinoma: evidence for histogenetic divergence of the enteric and nonenteric phenotypes. Hum Pathol. 2003;34:1101-1107.

(13.) Bashir AA, Robinson RA, Benda JA, Smith RB. Sinonasal adenocarcinoma: immunohistochemical marking and expression of oncoproteins. Head Neck. 2003;25:763-771.

(14.) Cathro HP, Mills SE. Immunophenotypic differences between intestinal-type and low-grade papillary sinonasal adenocarcinomas: an immunohistochemical study of 22 cases utilizing CDX2 and MUC2. Am J Surg Pathol. 2004;28: 1026-1032.

(15.) Franchi A, Massi D, Palomba A, Biancalani M, Santucci M. CDX-2, cytokeratin 7 and cytokeratin 20 immunohistochemical expression in the differential diagnosis of primary adenocarcinomas of the sinonasal tract. Virchows Arch. 2004;445:63-67.

(16.) Kennedy MT, Jordan RC, Berean KW, Perez-Ordonez B. Expression pattern of CK7, CK20, CDX-2, and villin in intestinal-type sinonasal adenocarcinoma. J Clin Pathol. 2004;57:932-937.

Accepted for publication August 11, 2006.

John A. Ozolek, MD; E. Leon Barnes, MD; Jennifer L. Hunt, MD

From the Department of Pathology, Children's Hospital of Pittsburgh (Dr Ozolek), and the Department of Pathology, University of Pittsburgh (Drs Barnes and Hunt), Pittsburgh, Pa. Dr Hunt is now with the Department of Pathology, Head/Neck/Endocrine Pathology, The Cleveland Clinic, Cleveland, Ohio.

The authors have no relevant financial interest in the products or companies described in this article.

This study was awarded Best Poster Presentation in Otorhinolaryngology Pathology at the United States and Canadian Academy of Pathology Annual Meeting in San Antonio, Tex, February 28, 2005.

Reprints: John A. Ozolek, MD, Department of Pathology, Children's Hospital of Pittsburgh, 3705 Fifth Ave, Pittsburgh, PA 15213 (e-mail: [e-mail]).

Table 1. Antibodies Used for Immunohistochemistry

                                                Antigen Retrieval/
Antibody     Dilution     Company               Pretreatment

CK7          1:100        Dako North America,   Microwave power 8,
                            Inc, Carpinteria,     2 x 5 min in
                            Calif                 citrate buffer
CK20         1:20         Dako                  Microwave power 8,
                                                  2 x 5 min in
                                                  citrate buffer
CDX-2        1:200        BioGenex              Protease digestion,
                            Laboratories,       5 min
                            San Ramon, Calif
Ki-67        1:25         Dako                  Microwave power 8,
                                                  2 x 5 min in
                                                  citrate buffer
34[beta]12   Prediluted   Dako (antibody to     2% pepsin, 3[degrees],
                            Moll types 1, 5,    10 min
                            10, 14)
p63          1:200        Imgenex, Inc, San     Microwave power 8,
                          Diego, Calif            1 x 5 min in EDTA
SMA          1:50         Dako                  None
S100         1:250        Novocastra            Citrate buffer
                            Laboratories,
                            Ltd, Newcastle
                            upon Tyne, United
                            Kingdom
Calponin     1:50         Dako                  None

Table 2. Clinical and Demographic Data *

                      Age, y
                                        Sex
Lesion    n    Mean   Median   Range   (M/F)

REAH     16     62      62     33-86   10/6
IP       10     58      68     17-81    7/3
SNAC     10     60      62     38-72    6/4
CS       10     51      45     31-86    8/2

* REAH indicates respiratory epithelial adenomatoid hamartoma; IP,
inverted papilloma; SNAC, sinonasal adenocarcinoma; and CS, chronic
sinusitis.

Table 3. Comparison of Basal/Myoepithelial Cell Markers in Sinonasal
Lesions *

Stain                           REAH                IP

S100 ([dagger])                 0/16                0/10
                                (0)                 (0)
SMA                             0/16                0/10
                                (0)                 (0)
34[beta]El2 ([double dagger])   16/16               10/10
                                (100) ([section])   100 ([parallel])
p63 ([paragraph])               16/16               10/10
                                (100) ([section])   100 ([parallel])
Calponin                        0/16                0/10
                                (0)                 (0)

Stain                           SNAC              CS

S100 ([dagger])                 3/10              0/10
                                (30)              (0)
SMA                             0/10              0/10
                                (0)               (0)
34[beta]El2 ([double dagger])   5/10              10/10
                                50 ([parallel])   (100) ([section])
p63 ([paragraph])               3/10              10/10
                                30 ([parallel])   (100) ([section])
Calponin                        0/9               0/10
                                (0)               (0)

* Values are number positive/total number (%). REAH indicates
respiratory epithelial adenomatoid hamartoma; IP, inverted papilloma;
SNAC, sinonasal adenocarcinoma; CS, chronic sinusitis; and SMA,
smooth muscle actin.

([dagger]) SNAC versus REAH for S100, P = .05.

([double dagger]) SNAC versus IP for 34[beta]E12, P = .03.

([section]) Cases with primarily basal cell staining.

([parallel]) Cases with primarily epithelial cell staining.

([paragraph]) SNAC versus IP for p63, P = .003.

Table 4. Comparison of Immunophenotype and Cytokeratin (CK) 7, CK20,
CDX-2, and Ki-67 Labeling Index (KLI) in Sinonasal Adenocarcinoma
(SNAC) *

                                             P Value
Stain         SNAC              REAH        ([dagger])         IP

CK7        9/10 (90)         16/16000)         .38         10/10000)
CK20       5/10 (50)          0/16(0)          .003         0/10(0)
CDX-2      3/10 (30)          0/16(0)          .05          0/10(0)
KLI,    12.8[+ or -]16.4   0.3[+ or -]0.5      .002      8.3[+ or -]3.1
%

Stain   P Value         CS         P Value

CK7      >.99       10/10000)       >.99
CK20      .03        0/10(0)         .03
CDX-2     .21        0/10(0)         .21
KLI,      .49     4.7[+ or -]6.5     .24
%

* Values are given for comparison of SNAC to REAH, IP, and CS,
respectively.

([dagger]) P Values are given for comparison of SNAC to REAH, IP, and
CS, respectively.

Table 5. Cytokeratin (CK) and CDX-2 Expression in Sinonasal
Adenocarcinoma (Intestinal-Type Adenocarcinoma [[TAC] and
Nonintestinal-Type Adenocarcinoma [non-[TAC]) *

Histologic Type                            CK7          CK20

Well to moderately differentiated (ITAC)   2-3+         N
Moderately differentiated (ITAC)           2-3+         Focal 1+
Well differentiated (ITAC)                 2-3+         N
Mucinous (ITAC)                            1-3+         2-3+
Moderately differentiated papillary        Focal 1-2+   2-3+
  mucinous (ITAC)
Moderately differentiated mucinous         2-3+         Focal 2-3+
  (non-ITAC)
Low-grade papillary (non-ITAC)             2-3+         N
Moderately to poorly differentiated        1-3+         N
  mucinous (non-ITAC)
Poorly differentiated adenocarcinoma       2-3+         N
  (non-ITAC)
Poorly differentiated adenocarcinoma       N            N
(non-ITAC)

Histologic Type                            34[beta]E12   CDX-2

Well to moderately differentiated (ITAC)   Rate+         N
Moderately differentiated (ITAC)           1-3+          Focal 1+
Well differentiated (ITAC)                 N             N
Mucinous (ITAC)                            N             3+
Moderately differentiated papillary        N             3+
  mucinous (ITAC)
Moderately differentiated mucinous         3+            N
  (non-ITAC)
Low-grade papillary (non-ITAC)             1-3+          N
Moderately to poorly differentiated        N             N
  mucinous (non-ITAC)
Poorly differentiated adenocarcinoma       1-3+          N
  (non-ITAC)
Poorly differentiated adenocarcinoma       3+            N
(non-ITAC)

* N indicates no staining; 1 +, weak intensity staining; 2+, moderate
intensity; and 3+, strong intensity.