Cytokeratin7 and cytokeratin19 expression in high grade cervical intraepithelial neoplasm and squamous cell carcinoma and their possible association in cervical carcinogenesis

The list of CIN and SCC was obtained from archives in the Department of Pathology at Nowon Eulji Medical Center, Eulji University. We retrieved 25 CIN3 and 30 SCC cases collected by cone biopsy, loop electrosurgical excision procedure (LEEP), or hysterectomy. The age of all CIN3 and SCC patients ranged from 27 to 79 years and the mean was 48 years. The stage of 30 SCC patients was classified into 11 cases of stage Ia, 10 stage Ib, 5 stage II, 3 stage III, and 1 stage IV according to the International Federation of Gynecology and Obstetrics staging. Ten non-neoplastic cervical specimens from the uteri removed for non-cervical pathologic conditions were also included. After reviewing all hematoxylin and eosin (HE) stained slides from each case, representative blocks for each case were selected. In two cases of CIN3 (LEEP) and one case of SCC (LEEP), immunohistochemical evaluation was not available in the slides made from initially selected blocks by repeat cutting due to technical reasons and other blocks for the same case were used for immunohistochemistry. This study was performed with the approval of the Institutional Review Board of Nowon Eulji Medical Center.

Immunohistochemical staining was performed using Dako Autostainer (DakoCytomation, Carpinteria, CA, USA). Four micron tissue sections were cut from selected blocks and positioned on poly-L-lysine coated slides. After deparaffinization and rehydration, antigen retrieval was performed using citrate buffer (pH 6.0) at 121 °C for 10 min. Endogenous peroxidase activity was blocked by 3% hydrogen peroxide for 5 min. The primary antibodies used in this study were CK7 (Dako, Carpinteria, CA, 1:100), CK19 (Dako, 1:100), and p16 (Dako p16INK4a kit). Color was developed with diaminobenzidine, and the slides were counterstained with hematoxylin.

CK7 and CK19 staining was interpreted as positive when approximately five to six contiguous cells showed cytoplasmic and/or membranous block staining [6]. With p16, nuclear or nuclear and cytoplasmic block staining was considered positive [18]. The immunoreactivity of CK7, CK19, and p16 was determined by assessing the staining intensity and percentage of stained cells. The staining intensity was rated as weak, moderate, and strong. The percentage of positive cells was scored as follows: negative – staining in less than 1% of cells, patchy – staining in 1% to 40% of cells, and diffuse – staining over more than 40% of cells.

In situ hybridization (ISH) was performed using the INFORM HPV III Family 16 Probe (B) (Ventana Medical Systems, Tucson, AZ, USA), according to the manufacturer’s recommendations. The probe cocktail has demonstrated affinity for the following HR HPV genotypes: 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, and 66. The HPV signal patterns were classified as the episomal or integrated form, found within the nuclei of cervical epithelial cells. The episomal HPV was defined as large globular dense nuclear staining. The integrated HPV was defined as discrete dot-like signals or scattered tiny particles in the nuclei.

In the non-neoplastic cervix (Fig. 1a), CK7 staining was strong in the columnar cells at the SCJ, faintly seen in the cells at the transformation zone (Fig. 1b), and negative in squamous cells at the ectocervix. CK19 was weakly stained in the columnar cells at the SCJ, squamous cells at the transformation zone (Fig. 1c), and basal layer cells at the ectocervix.

The results of immunohistochemistry for CK7, CK19, and p16 in CIN3 are shown in Fig. 2 and their detailed expression patterns and HPV status are demonstrated in Fig. 3. The expression of CK7, CK19, and p16 was positive in all of 25 CIN3 cases. There were 12 (48%) patchy stained cases of CK7, and 10 (40%) of CK19, topographically distinct. Patchy CK7 expression was usually seen in the upper layer of CIN3, most intensely in the superficial layer and gradually weaker in the lower layer. In some cases of CIN3 located in the ectocervix (CIN3#12, Fig. 4a), CK7 staining was strong in the cells in the upper layer (Fig. 4d). Conversely, patchy CK19 expression was most intense in the basal layer of CIN3 and gradually weaker toward the upper layer (Fig. 4g). Diffuse staining of CK7 was seen in 13 (52%) and 15 (60%) cases of CK19, with homogeneous expression in entire layer of CIN3 located at the transformation zone (CIN3#21, Fig. 4b, e and h) and the SCJ (CIN3#23, Fig. 4c, f and i). p16 expression was diffuse in all CIN3 cases with moderate (Fig. 4j and l) to strong intensity (Fig. 4k).

HPV was detected in all CIN3 cases including 11 (44%) cases with combined episomal and integrated form and 14 (56%) cases with integrated form. Episomal HPV DNA was usually seen in the upper layer corresponding with CK7 staining area (Fig. 4m and o), while integrated HPV DNA was noted in entire layer (Fig. 4n).

The results of immunohistochemistry for CK7, CK19, and p16 in SCC are shown in Fig. 5 and their detailed expression patterns and HPV status are demonstrated in Fig. 6. While the expression of CK19 and p16 was positive in all 30 SCC cases, CK7 expression was positive in 20 cases (67%) and negative in 10 cases (33%). Patchy staining of CK7 was seen in 13 (43%) and 19 cases (63%) of CK19. Of the typical SCCs (13 cases, 43%) showing solid tumor nests with/without keratinous pearl (SCC#8, Fig. 7a), CK7 staining was positve in 8 cases (62%) and negative (Fig. 7d) in 5 cases while CK19 staining was positive (Fig. 7g) in all cases. Interestingly, in the case of SCC#28, CK7 staining was noted in a keratinizing tumor nest. Topologically distinctive staining pattern of CK7 and CK19 was seen in SCC with central cystic space within the tumor nest, such as SCC#19. In SCC#19 (Fig. 7b), CK7 staining was diffusely positive in overall tumor nest but stronger in the cells of inner portion of the nest and cellular debris shedding within the space (Fig. 7e). However, CK19 positive cells were patchily present along the periphery of the invasive tumor nests (Fig. 7h) (Addittional File 1: Figure S1). The central cystic space within the tumor nest was seen in 14 cases (47%) and 10 of 14 cases (71%) were positive for CK7. In SCC showing infiltrating cord-like or branching growth pattern (Fig. 7c), CK7 and CK19 staining was both diffuse and strong in the tumor cells, and their staining was intensified in the invasive front and tumor cells with glandular differentiation (SCC#29, Fig. 7f and i). This growth pattern was seen in 3 cases (10%) and 2 of 3 cases (67%) were positive for CK7. Despite strong CK7 and CK19 expression in a subset of SCC, overall staining of CK7 and CK19 in SCC was weaker than that in CIN3

The expression of p16 was also weaker in SCC compared with CIN3, with even one patchy stained case (SCC#20, Fig. 6). In SCC#19, the staining of p16 was diffuse but disappeared toward the inner portion of the nest (Fig. 7k). It was notable that in SCC with diffuse and strong CK7 staining, p16 staining was weak (SCC#29, Fig. 7l), whereas in SCC with no CK7 staining, p16 staining was strong (SCC#8, Fig. 7j), suggesting an inverse correlation between CK7 and p16 in SCC, although not consistent in all cases.

HPV was positive in all SCC samples, displaying mixed episomal and integrated form in 14 (47%) and integrated form in 16 (53%) of 30 SCC cases. The cases with integrated HPV DNA revealed no topographic distinction. In SCC#19, an episomal form was seen in the entire tumor nest but more frequently in central portion of the nest, overlapping with the expression of CK7 (Fig. 7n). However, an episomal HPV was not always concordant with CK7 expression because HPV with the integrated form only was also seen in tumor cells with strong CK7 positivity such as SCC#29 (Fig. 7o). Inversely, in CK7 negative case such as SCC#8, an episomal form of HPV DNA was also noted (Fig. 7m).