Clinical and histopathologic prognostic implications of

Archives of Oral Biology 99 (2019) 1–8

Contents lists available at ScienceDirect

Archives of Oral Biology journal homepage: www.elsevier.com/locate/archoralbio

Clinical and histopathologic prognostic implications of the expression of cytokeratins 8, 10, 13, 14, 16, 18 and 19 in oral and oropharyngeal squamous cell carcinoma

T

⁎

Rima A. Safadia,b, ,1, Niveen I. Abdullahc, Rolla F. Alaarajc, Dima H. Baderd, Darshan D. Divakare, Abed A. Hamashaa,b,1, Maher A. Sughayerc a

Faculty of Dentistry, Jordan University of Science and Technology, Irbid, Jordan In sabbatical leave to College of Dentistry, King Saud Bin Abdul Aziz University for Health Sciences Riyadh, Saudi Arabia King Hussein Cancer Center, Amman, Jordan d Ministry of Health, Amman, Jordan e College of Applied Medical Science, King Saud University, Riyadh, Saudi Arabia b c

A R T I C LE I N FO

A B S T R A C T

Keywords: Cytokeratin Immunohistochemistry Oral Prognosis SCC Survival

Objectives: To identify cytokeratins (CK) of significant correlations with clinical and histopathologic prognostic parameters in oral and oropharyngeal squamous cell carcinoma (SCC). Design: The sample consisted of 100 cases retrieved from the archives of the Pathology Department/ King Hussein Cancer Center/Amman/ Jordan. Recorded data included: age, gender, location, grade, depth of invasion, the presence of epithelial dysplasia, tumor size, lymph node metastasis, number of positive lymph nodes, distant metastases, clinical stage, local recurrence, treatment modalities and 5-year survival rate. Immunohistochemical staining of 7 cytokeratins: 8, 10, 13, 14, 16, 18, and 19 was performed using standard protocols. Stained sections were digitized and analyzed using ImageJ-color deconvolution to identify the percentage of cytokeratin-positive area (score). Statistical tests used were: student t-test, analysis of variance, bivariate analysis and logistic regression. Results: Lower CK8,18, 19 scores correlated with lower 5-year survival rate. Higher CK19 and lower CK 10, 14, 16 scores were associated with distant metastasis. Increased CK8, 18, 19 scores correlated with higher stage and with higher depth of invasion. Increased CK18 scores correlated with increased local recurrence. Higher CK10, 13, 16 scores correlated with well-differentiated grade. Higher CK19 and lower CK16 scores were associated with adjacent epithelial dysplasia. Regression analysis showed that better 5-year survival rate was significantly correlated with increased CK16, decreased CK18 and 19 scores. Conclusion: Expression scores of a panel of cytokeratin are potential prognostic indicators for 5-year survival and correlates with other prognostic parameters.

1. Introduction Oral squamous cell carcinoma (OSCC) is the most common malignant tumor of the oral and maxillofacial region with a 5-year-survival of approximately 50–60% (Coelho et al., 2015; Zhong, Chen, Zhang, & Zhang, 2007). It comprises about 95% of different cancers affecting the oral cavity (Krishnan et al., 2016). According to the World Health Organization (WHO)'s International Classification of Diseases, “Oral cancer” may be defined as any malignant neoplasm occurring on the lips (both vermillion border and oral aspect) and within the oral cavity

(which includes the anterior 2/3 of the tongue, buccal and labial mucosa, gingiva, hard palate, retromolar pad and floor of the mouth). Oral and pharyngeal cancer, grouped together, is the sixth most common cancer in the world (Moore, Pierce, & Wilson, 2000; Warnakulasuriya, 2009). The 5-year survival rate of OSCC cancer patients did not show significant improvement despite the advances in treatment modalities, which makes prevention and early diagnosis of OSCC as the most significant factors to improve the survival rates. The aggressiveness of this tumor is related to different factors, including clinical stage,

⁎

Corresponding author at: Faculty of Dentistry, Jordan University of Science and Technology, P.O Box: 3030, Irbid 22110, Jordan. E-mail addresses: [email protected], [email protected] (R.A. Safadi). 1 Current address: College of Dentistry, King Saud bin Abdul Aziz University for Health Sciences P.O Box: 22490, Mail code: 1243, Riyadh 11426, Saudi Arabia. https://doi.org/10.1016/j.archoralbio.2018.12.007 Received 4 October 2018; Received in revised form 13 December 2018; Accepted 14 December 2018 0003-9969/ © 2018 Elsevier Ltd. All rights reserved.


R.A. Safadi et al.

plugged in ImageJ image analysis system. ImageJ is a public domain Java image processing program developed at the National Institute of Health (Bethesda, Maryland, USA). A detailed description of color deconvolution analysis was published previously (Ruifrok & Johnston, 2001). For each case, all digitized images were analyzed for positive epithelial area divided by the total epithelial area in the analyzed field to calculate the percentage of positive cytokeratin expression. For each case, the average of the positive area percentage in all analyzed fields was considered as the score for the respective cytokeratin.

histological grade, and depth of invasion, tumor size and anatomical location. Other factors that correlate with prognosis and survival of OSCC patients including the level of involvement of neighboring tissue, the involvement of regional lymph nodes, and distant metastasis (Bryne, 1998; Dantas da Silveira et al., 2007). The expression of cytokeratins (CK) and cell cycle proteins have been investigated to better understand carcinogenesis. Cytokeratins are a family of intermediate filaments that play an essential role in the structural integrity and stability of the epithelial cell (Coelho et al., 2015). Pathologically, when the cell growth rate and/or level of differentiation are changed, the expression patterns of the 20 different subtypes may be altered. The knowledge of the prognostic significance of CK expression patterns in SCC is not well explored yet (van der Velden, Manni, Ramaekers, & Kuijpers, 1999). The present study aimed at investigating and correlating the immunohistochemical expression of a panel of cytokeratins with the clinical and histopathologic prognostic parameters of oral and oropharyngeal squamous cell carcinoma in order to identify the cytokeratin types most significantly correlating with an overall survival rate.

2.4. Statistical analysis Data were entered into the statistical package for social science program, version 22, (SPSS Inc; Chicago, IL). Statistical tests used where: student t-test, analysis of variance (ANOVA), bivariate analysis and logistic regression. The study cases were categorized according to the location into three categories: intraoral, oropharyngeal and lip. Furthermore, the cases were grouped into two categories according to the depth of invasion. In order not to group the cases arbitrarily, the cutoff point that gave the most significant correlation of cytokeratin scores with the depth of invasion was 1.25 cm (Roc curve, area under the curve). Lymph node involvement was considered as a continuous variable and as a categorical (yes, no) variable. The depth of invasion was analyzed as a categorical variable (above or below 1.25 cm). According to the histopathologic grade, tumors were categorized into 3 groups: well differentiated, moderately differentiated and poorly differentiated.

2. Materials and methods 2.1. Case retrieval The archives of the Pathology Department in King Hussein Cancer Center, Amman, Jordan were searched over a 10-year period for cases diagnosed histopathologically as squamous cell carcinoma of the oral cavity, oropharynx, paranasal sinuses and lips vermilion borders. Identified patient records and data were retrieved. Recorded information included: age, gender, tumor location, histopathologic grade, depth of invasion, presence of epithelial dysplasia in association with the tumor, tumor size, presence of lymph node metastasis and number of positive lymph nodes, presence of distant metastases, clinical stage, local recurrence after surgical resection, treatment modalities and 5year survival rate. The clinical staging was done according to the American Joint Committee on Cancer tumor-node-metastasis staging system (Edge & Compton, 2010). The respective micro slides were retrieved and evaluated for the histopathologic information. Paraffin blocks were evaluated for the amount of remaining tissues and appropriateness for further recuts.

3. Results The sample consisted of 100 cases with 74 males and 26 females. The average age was 58.3 years ranging from 24 to 102 years being 57.6 for males and 60 for females. Tumors distribution by location was as follows: tongue: 39%, lower lip: 15%, buccal mucosa 11%, oropharynx: 10%, floor of the mouth: 7%, palate 6%, gingiva 6%, retromolar area 4%, upper lip 2%. Tumors distribution by the main anatomic area was: Intraoral: 73%, Lip 17%, oropharyngeal 10%. No squamous cell carcinoma cases of the paranasal sinuses were identified in our cohort. Tumor size ranged from 0.2 cm to 6.5 cm with an average of 1.8 cm and a median of 1.5 cm. Data related to lymph node involvement assessed by microscopic examination was available for 47% of the cases. According to the presence of positive lymph node at the time of diagnosis, 25%% of the patients had positive lymph nodes that ranged in number from 1 to 27. Distant metastasis was present in 19 patients as indicated in the patients’ medical records to be assessed using imageanalysis while 81 were free of distant metastasis at time of diagnosis. Study patients had their TNM stage distributed as follows: stage I: 14, stage II: 13, stage III: 12, stage IV: 61. The cases were grouped by stage for statistical analyses into 2 groups: group 1 consisted of stage I – III (39 cases), and group 2 consists of stage IV (61 cases). The grouping was done since the number of cases in each of the first 3 stages was low for statistical significance. Group 1 consisted of stage I – III (39 cases), and group 2 consisted of stage IV (61 cases). When the staging was encoded into 2 groups 3 CKs turned to show significant correlation. However, regression analysis showed none to keep its significance. Most frequent treatment modality was a combination of surgery, chemotherapy, and radiotherapy (53 out of 100) while 2-modality treatment was noted in 40 patients: 20 had surgery and radiotherapy while 20 had chemotherapy and radiotherapy. Five patients had radiotherapy alone and 2 had no further treatment. Patients who had local recurrence after surgical resection were 28. The 5-year survival rate information was available for all patients: 52 survived the first 5 years and 48 did not. According to tumor

2.2. Immunohistochemical staining The paraffin-embedded tissue had seven sections of 4 μm- thickness cut and mounted on vectabond-coated and charged glass micro slides (DiaPath, SuperFrost Plus, 060SFP, Italy). One additional tissue section was cut for H&E staining. Immunohistochemical staining was performed following the standard procedure using DakoEnVision™+ Dual Link System-HRP, visualized using diaminobenzidine solution (Dako, Glostrup, Denmark) and counterstained with Meyer's hematoxylin using automated stainer. All primary antibodies were diluted at 1:150 which was the optimum dilution ratio used at room temperature for 1hour incubation period. The primary antibodies used from Dako, Glostrup, Denmark included: Mouse Monoclonal Anti-keratin 8, 10, 13, 14, 16, 18 and 19. One tissue section in each immunohistochemical run was incubated with non-specific mouse IgG (Biogenex, California, USA) at 1:150 dilutions, replacing the primary antibody as a negative control. No additional sections were used for positive control as the study sections had internal positive control in each run. 2.3. Image analysis For each immunostained section, 3 representative fields were digitized at 100X magnification for purpose of image analysis. Digitized images were analyzed using color deconvolution software 2


0.29 0.59 0.25 0.85 0.84 0.04 0.49

No (72)

3.6 10.2 15.8 30.8 13.1 5.4 3.1 5.0 11.7 19.1 30.0 13.7 9.8 2.1

3.1. CK 8, 10, 13, 14, 16, 18 and 19 expression correlations with clinical prognostic parameters

0.08 0.31 0.59 0.78 0.6 0.07 0.03 3.6 11.4 16.1 28.2 12.5 6.2 3.6 5.7 8.6 17.2 34.0 13.8 8.7 1.7 0.75 0.08 0.34 0.06 0.06 0.91 0.001 3.6 11.7 17.6 32.5 14.2 7.8 1.6 4.1 7.5 14.7 24.8 9.3 6.6 6.6 0.06 0.5 0.17 0.52 0.79 0.02 0.5 8.5 11.5 21.2 28.5 13.5 13.1 5.2 0.27 0.01 0.01 0.09 0.01 0.80 0.25

No (49) Yes (17) P value Lip (17)

2.8 18.7 17.5 37.7 19.4 5.4 0.4 4.3 9.4 17.7 30.2 12.8 7.6 3.3

7.9 4.8 4.9 21.2 4.6 7.0 3.4

3.2. CK 8, 10, 13, 14, 16, 18 and 19 expression correlations with histopathologic prognostic parameters

Intraoral (73)

Oropha ryngeal (10)

Statistical correlation of cytokeratins` score with different tumor clinical parameters is presented in (Table 1). ANOVA test showed that CK10, CK13, CK16 scores were statistically significantly less expressed in oropharyngeal SCC compared to intraoral and lower lip SCC. Fig. 1 illustrates an oropharyngeal SCC case which is negative for CK10 and 13 but positive for CK18. Decreased expression of CK 18 and 19 was significantly associated with mortality within the first 5-years (P < 0.05, student t-test). CK8 decreased expression is also noted in those who did not show 5-yearsurvival, however, this did not reach statistical significance (P = 0.06, student t-test). Increased CK19 expression was statistically strongly associated with distant metastasis at the time of diagnosis (P < 0.001, student t-test). On the other hand, decreased expression of CKs 10, 14, 16 was noted in patients who presented with distant metastasis at the time of diagnosis, but P values were close to statistical significance. (P = 0.06, 0.06, 0.08 respectively, student t-test). Increased CK19 expression was significantly correlated with higher clinical stage. Increased CK8 and CK18 expression showed a similar trend of association with higher clinical staging with a p-value close to but not reaching statistical significance (P = 0.08, 0.07 respectively, student t-test) (Fig. 2). Tumors with higher CK18 score were more prone to local recurrence (P = 0.04, student t-test) after surgical resection. Other clinical prognostic parameters included the number of positive lymph nodes and patient gender. The study showed no statistical significant correlation between the numbers of positive lymph nodes with any CK score. The correlation coefficients ranged from 0.06 to 0.11 for all study cytokeratins using bivariate analysis. Tumor size similarly did not show any significant statistical correlation with the expression of CKs (correlation coefficients ranged from 0.03-0.2 for all CKs, Bivariate analysis). In addition, CKs scores did not show statistical correlation with patient gender (p-value ranged from 0.3- 0.7 for all CKs, student t-test).

5.0 10.8 14.1 28.5 10.8 6.1 2.4

Stage I-III (39) No (81) Yes (19)

Distant metastasis

Student ttest P value 5 year survival ANOVA

Table 2 presents the statistical correlations of the study cytokeratins` scores with the different histopathologic prognostic parameters investigated in this study. CK10, 13, 16 showed statistically significant increased expression in well-differentiated tumors (P = 0.04, 0.01, < 0.001, ANOVA) (Fig. 3). CK18 and CK19 scores were statistically significantly higher in tumors with depth of invasion of > 1.25 cm (Fig. 4) (P = 0.02 and < 0.001 respectively, student t-test).CK8 also showed increased expression in tumors with more depth of invasion with a p-value close to but not reaching statistical significance (P = 0.06, student t-test). In SCCs associated with epithelial dysplasia, CK19 showed significantly higher scores (Fig. 5) (P = 0.01, student ttest) while CK16 showed significantly lower scores (P = 0.05, Student t-test). 3.3. Regression analysis results

CK8 CK10 CK13 CK14 CK16 CK18 CK19

Tumor location

Clinical parameter/statistical test CK type/score Category (N)

Table 1 The correlation of cytokeratins 8, 10, 13, 14, 16, 18, 19 expression scores with clinical prognostic parameters.

Student ttest P value

TNM stage

Stage IV (61)


Yes (28)

histopathologic grade, 27 cases were well differentiated, 50 moderately differentiated and 23 poorly differentiated. Presence of epithelial dysplasia in association with the tumor was noted in 44 cases. The depth of invasion measurement was available for 96 cases out of 100. Depth of invasion ranged from 0.1 mm to 4.5 mm with 80 cases being less than 1.25 mm and only 16 were more than 1.25 mm. The mean Depth of invasion was 0.86 mm and the median 0.7 mm.

Local recurrence


R.A. Safadi et al.

Logistic regression analysis was conducted for all variables that showed a significant correlation with the survival rate. The regression 3


R.A. Safadi et al.

Fig. 1. 69-year-old male with poorly differentiated SCC of the oropharynx. (A): Hematoxyline and eosin stain (4x), (B): CK18 positive immunostain (score = 51%) (4x), (C): CK10 negative immunostain (score = 0%) (4x), (D): CK 13 negative immunostain (score = 0%) (10x).

study came after searching the scientific literature for potentially clinically significant types. Cytokeratins` 8/18 expression had been shown to correlate with early invasion of OSCC (Fillies et al., 2007), CK10 and CK13 have been suggested as markers of dysplastic epithelium, and have been used to assess the malignant progression in premalignant lesions (Garcia et al., 2016). Cytokeratin16 expression occurred in the suprabasal keratinocytes of hyperproliferative conditions as well as in cells with high turnover rates. Its expression is considered as a marker of proliferation in keratinocytes (Fonseca & Do Carmo, 2001). CK19 expression in the suprabasal layer of oral mucosa correlated with premalignant dysplastic changes in oral epithelium (Safadi, Musleh, Al-Khateeb, & Hamasha, 2010). However, the clinical significance of CK19 detection in OSCC has not been confirmed yet (Zhong et al., 2007). In the current study, our 100 cases were categorized according to their location into 3 categories: intraoral (73 cases/ 73%),

model showed that the variables that retained their significant correlation with 5-year survival rate after adjustment for confounders to be: CK16, CK18, CK19, and tumor location (Table 3). Tumors with increased expression of CK16 or lower expression of CK18 or lower expression of CK19 showed a better chance for 5-year survival. Tumors occurring in oropharynx had the lower expectation for 5-year survival while lower lip tumors had the best prognosis among the 3 locations categories. 4. Discussion The present study aimed at quantitively correlating the immunohistochemical expression of cytokeratins 8, 10, 13, 14, 16, 18 and 19 with clinical and histopathologic prognostic parameters in order to identify the cytokeratin types most significantly correlating with overall survival rate. The selection of the cytokeratin panel included in this

Fig. 2. 62-year-old female with poorly differentiated SCC of the mandibular gingiva, positive for distant metastasis at time of diagnosis and passed away within first 5 years. (A): Hematoxyline and eosin stain (4x), (B): CK 19 positive immunostain (score = 47%) (4x), (C): CK 18 positive immunostain (score = 49%) (4x). 4


R.A. Safadi et al.

Table 2 The correlation of cytokeratins 8, 10, 13, 14, 16, 18, 19 scores with histopathologic prognostic parameters. Histopathologic parameter/ statistical test

Histopathologic Grade

CK type/score Category (N)

Well differentiated (26)

Moderately differentiated (51)

CK8 CK10 CK13 CK14 CK16 CK18 CK19

4.2 14.1 22.4 33.4 17.9 9.0 3.1

4.7 11.2 13.5 9.3 15.0 4.8 2.5

ANOVA

Depth of Invasion (cm)

Student ttest

Adjacent epithelial dysplasia

Poorly Differentiated (23)

P value

< 1.25 (80)

> 1.25 (16)

P value

Yes (44)

No (56)

4.8 4.9 16.0 25.4 2.6 9.3 3.5

0.96 0.04 0.01 0.28 < 0.001 0.16 0.83

4.3 9.4 17.4 31.2 13.4 7.0 2.8

5.7 16.1 12.8 25.4 11.4 7.6 3.2

0.06 0.5 0.17 0.52 0.79 0.02 0

4.1 11.5 16.4 28.1 9.9 6.5 3.8

4.4 12.0 16.4 53.0 15.6 5.4 0.5

Student ttest

0.87 0.89 0.99 0.14 0.05 0.61 0.01

patients in our study had a higher age average compared to male, which comes in accordance with a previous study of Stell et al. who found when analyzing a series of 842 OSCC patients that women were older at presentation than men (Stell, 1992). There was no statistical correlation between CKs expression scores with patient gender in our study which was in concordance with the study done by Kaliszewski et al. where CK19 expression did not associate with gender (Kaliszewski et al., 2016). In the current study, tumor size ranged from 0.2 cm to 6.5 cm with an average of 1.8 cm. In another study done in Rio de Janeiro, Brazil, most OSCC had tumor size of up to 4 cm in their greatest diameter, exhibiting a mean of 3.4 cm ranging from 0.2 to 12 cm (Pires et al., 2013). CKs expression scores did not show significant statistical correlations with tumor size. Son E-M et.al, observed that pancreatic neuroendocrine tumors expressing CK19 were related with higher WHO grade, larger tumor size, lymphovascular and perineural invasions, and higher clinical stage (Son et al., 2015). Another study was in line with Son E-M et al. showing CK19-negative tumors tending to be smaller than the typical tumor size in the series (2.5 vs. 3.6 cm) (Ali, Serra, Asa, & Chetty, 2006). It is noteworthy that the latter 2 studies were concerned with the presence and absence of CK expression rather than CK expression scores, which might have contributed to variation in results. Further studies are recommended. At the time of their diagnoses, 47% of our patients presented with positive lymph node involvement. A study of 347 Taiwanese patients diagnosed with OSCC showed a lower percentage (32.3%) of patients to be associated with positive lymph nodes (C.-C. Lee et al., 2015). Differences may be attributed to delay in diagnosis or variances in the predisposing factor of the disease among different populations. Our study showed no statistical significant correlation between the CK expression scores with the numbers of positive lymph nodes. In a previous study done by Han X et al. regional lymph nodes metastases, and advanced clinical stage was significantly associated with the positive CK19 expression (P < 0.05) (Han, Zhao, Ji, Xu, & Lou, 2013). The variance from our study could be attributed to the different

oropharyngeal (10 cases/ 10%) and lip (17 cases /17%). A similar trend of SCC site-prevalence has been observed in a Jamaican study of 384 OSCC, which found that 79% to be distributed intraorally, 17% involved the oropharynx and 4% the lips. Likewise, the distribution in a Canadian study was 70% in the intraoral, 17% oropharyngeal, and 13% lip cancer. Additionally, in a study of oral cancer in a Chinese population on a cohort of 1240 patients; the site distribution was found to be 93.43% intraorally, 3.4% involving the lips and 3.17% involving the oropharynx (Liu et al., 2017; Nyi, Gibson, Hanchard, Waugh, & McNaughton, 2014). In our study, the lateral border of the tongue was found to be the most commonly affected site followed by lower lip, buccal mucosa and oropharynx. Variable but comparable results have been identified in the literature. In Iran, the tongue and the buccal mucosa showed the highest prevalence, and in Hungary, Nemes et al. (2006) showed that the floor of mouth, lips and the tongue were the most common areas. Whilst the border of the tongue is regarded as the commonest site for OSCC in America and Europe, the buccal mucosa is the most common site for OSCC in Southeast Asia (Pires et al., 2013). On the other hand, in Thailand, it was described that 50% of the OSCC affected the alveolar ridge and gingiva while in Nigeria, the mandibular and maxillary gingiva were the sites most commonly affected, followed by the tongue. The differences in percentages of site prevalence among studies could be attributed to a multitude of factors including sociodemographic, race, oral habits and the significant difference in the number of cases (Effiom et al., 2008; Jainkittivong, Swasdison, Thangpisityotin, & Langlais, 2009; Maleki et al., 2015; Nemes, Boda, Redl, & Marton, 2006; Pires et al., 2013). Our sample included 74 males and 26 females with a male to female ratio of about 3:1 which is comparable with the incidence reported by the Surveillance, Epidemiology and End Results (SEER) program in the United States that showed a 2-3:1 ratio depending on race and age. A Canadian review of oral cancer reported the overall ratio of males to females to be 2:1 and 1:1 in patients under the age of 40 (Laronde, Hislop, Elwood, & Rosin, 2008; National Cancer Institute). Female

Fig. 3. 60-year-old male with well differentiated SCC of the retro molar region. (A): Hematoxyline and eosin stain (10x), (B): CK13 positive immunostain (score = 8%) (4x), (C): CK16 positive immunostain (score = 11%) (10x). 5


R.A. Safadi et al.

Fig. 4. 59-year-old male with moderately differentiated SCC of the lower lip. Depth of invasion was of more than 1.25 cm: (A): Hematoxyline and eosin stain (4x), (B): CK19 positive immunostain (score = 28%) (4x).

were absent (Bloor, Seddon, & Morgan, 2001), this comes in accordance with our finding. In another study, CK13 was also expressed in tumor nests of welldifferentiated l OSCC (Wang, Liu, & Chiang, 2006). Of all investigated CKs in our study, increased CK19 expression was statistically strongly associated with distant metastasis. CK19 has been related with the migration and infiltration of cancer cells and may be linked with the development of distant metastasis due to its stimulation of extracellular degradation and cell motility (Lee et al., 2016). On the other hand, decreased expression of CKs 14, 16, 10 was noted in patients who presented with distant metastasis at the time of diagnosis, however, P values were close to statistical significance. (Student t-test, P = 0.06, 0.06, 0.08). As noted in our study, higher CK10 and 16 scores were correlating with the well differentiated SCC which could imply an inverse relationship of CK 10 and 16 expressions with distant metastasis. No mention of the direct correlation of cytokeratin expression with distant metastasis was found in the published literature, only indirect inferences have been mentioned in some studies (Dantas da Silveira et al., 2007; Yoshida et al., 2015). Our results showed that tumors, which expressed more CK18, were more prone to local recurrence compared with those expressing lower levels. In a study by Son et al. CK18 and CK8 expression were recognized as predictor of progression and poor prognosis after curative resection in OSCC patients (Son et al., 2015). Depth of invasion has been noted to be an independent predictor of disease-specific survival, and thus recently has been incorporated into the American Joint Committee on Cancer T category to improve

methodology as our comparisons were based on digitized expression scores rather than objective quantification or nominal description. In our study, the presence of epithelial dysplasia in association with the tumor was noted in 44 (44%) cases. In a study done by GonzalezMoles et al. the results were close where 37 out of 74 cases (50%) presented dysplasia adjacent to OSCC (Gonzalez‐Moles, Ruiz‐Avila, Rodriguez‐Archilla, & Martinez‐Lara, 2000). Our results showed that CK19 expression score was significantly higher in tumors that were associated with adjacent epithelial dysplasia (student t-test, P = 0.01). A study conducted by Kale et al. concluded that CK19 was expressed in 70% of clinically normal appearing mucosa associated with adjacent tumors, some of which were severely dysplastic, suggesting that CK19, could be associated with field cancerization (Kale, Mane, Babji, & Gupta, 2012). However, no other studies addressing the correlation of cytokeratin expressions in epithelial dysplasia adjacent to SCC were identified. Our results revealed statistically significant association of increased expression of CK10, CK13, and CK16 with intraoral and lower lip squamous cell carcinoma compared with oropharyngeal SCC. We suggest that a study with a larger sample size to facilitates a better understanding of CK profile in SCC of different locations in the oral cavity and oropharynx. Wherein our study CK10, 13, 16 showed statistically significant increased expression in well-differentiated tumors (student t-test, P = 0.04, 0.01, < 0.001). Bloor et al. found that in tumor islands of well and moderately differentiated SCC the CK4/13 complex was coexpressed with CK1/10 but in poorly differentiated carcinomas they

Fig. 5. 58-year-old male patient with moderately differentiated oropharyngeal SCC associated with adjacent surface epithelial dysplasia. (A): Hematoxyline and eosin stain (10x), (B): CK19 positive immunostain (score = 6%) (10x). 6


R.A. Safadi et al.

Funding

Table 3 Final regression analysis model of the study variables that significantly associated with survival. Variables

CK16 score CK18 score CK19 score Location

Standard Error

0.04 0.03 0.08 0.03

Significance

0.02 0.04 0.02 0.02

Exponential B

0.91 0.94 0.84 0.53

This research was funded by Deanship of Scientific Research/Jordan University of Science and Technology, Irbid, Jordan. Grant number: 197/2009.

95% confidence interval for exponential B Lower

Upper

0.84 0.89 0.72 0.31

0.98 1.00 0.98 0.90

Conflicts of interest statement No potential conflicts of interest exist. Competing interests None declared.

discrimination among patient subgroups with respect to disease specific survival (Lydiatt et al., 2017). In our study, CK 8, 18 and 19 expression scores were statistically significantly associated with the tumors with more depth of invasion (in our parameters > 1.25 cm). This indicates that CK8, 18, 19 increased expressions in SCC could be of potential prognostic indicator especially in terms of the disease specific survival. Our cut off value could be different when other CKs are investigated since in our study, it was the value that gave the highest specificity and sensitivity with the study CKs scores (Roc curve). Further research in this regard is recommended. The increased CK19 expression in our result was significantly correlated with higher TNM stage. CK8 and CK18 expression showed a similar trend but not reaching the statistical significance. It has been reported that low-molecular-weight cytokeratins` expression (CK8/18 and 19) in SCC is associated with higher degree of cell cycle deregulation, thus more correlation with advanced clinical stage (Frohwitter et al., 2016; Makino et al., 2009; Nanda, Ranganathan, Devi, & Joshua, 2012). Our linear regression model showed that when CK16 expression was higher, the chance for 5-year survival is better while the expression of CK18 and 19 was inversely related with the 5- year survival. CK16 down-regulation has been observed to associate with distant metastasis and disease-specific deaths since its down-regulation may cause loss of desmosomal integrity and increased mitosis (Harris et al., 2014). In another study, positive CK19 staining in tongue SCC correlated with a worse disease-specific survival when compared with the CK19 negative tumors (Ernst et al., 2016). In addition, many published studies confirmed that moderately and poorly differentiated OSCC expressed more CK 19 than well-differentiated tumors, which in turn correlates with a poorer prognosis (Ebrahimi et al., 2014; Jiang, Gu, Moore-Medlin, Nathan, & Hutt-Fletcher, 2012; Safadi et al., 2010; Siriwardena, Tsunematsu, Qi, Ishimaru, & Kudo, 2018). Furthermore, the linear regression model retained the significant correlation of oropharynx as a location with lower expectation for 5year survival while lower lip tumors had the best prognosis among the 3 locations categories. This is in agreement with other published studies where the percentage of deaths was higher in squamous cell carcinoma diagnosed in the oropharynx than in an oral location (Le Campion et al., 2017; Varsha et al., 2015). It can be anticipated that squamous cell carcinoma of the posterior anatomic location can go unnoticed for a longer time, delaying the diagnosis. In addition, surgical inaccessibility and proximity of vital structures may be the contributing factor to the relatively poorer survival rates of oropharyngeal squamous cell carcinoma. In conclusion, the results of our study showed that cytokeratin’s immunohistochemical expression is potentially a valuable predictor of prognosis in patients with Oral/oropharyngeal squamous Cell Carcinoma. Based on the investigated cytokeratins` expression in our study, a better 5-year survival rate correlated with increased CK16 expression and decreased CK18, 19 expressions. The study results may be of support to new SCC treatment modalities that could improve the overall survival rate.

Ethical approval The Research proposal has been approved by IRB committees of both King Hussein Cancer Center (IRB number 09 KHCC 88) and Jordan University of Science and Technology (IRB number 197/2009). References Ali, A., Serra, S., Asa, S. L., & Chetty, R. (2006). The predictive value of CK19 and CD99 in pancreatic endocrine tumors. The American Journal of Surgical Pathology, 30(12), 1588–1594. Bloor, B., Seddon, S., & Morgan, P. (2001). Gene expression of differentiation-specific keratins in oral epithelial dysplasia and squamous cell carcinoma. Oral Oncology, 37(3), 251–261. Bryne, M. (1998). Is the invasive front of an oral carcinoma the most important area for prognostication? Oral Diseases, 4(2), 70–77. Coelho, B., Peterle, G., Santos, M., Agostini, L., Maia, L., Stur, E., ... Freitas, F. (2015). Keratins 17 and 19 expression as prognostic markers in oral squamous cell carcinoma. Genetics and Molecular Research:GMR, 14, 15123–15132. Dantas da Silveira, E. J., Pina Godoy, G., AlvesUchôa Lins, R. D., Silva Arruda, M.d. L., Formiga Ramos, C. C., de Almeida Freitas, R., ... Guedes Queiroz, L. M. (2007). Correlation of clinical, histological, and cytokeratin profiles of squamous cell carcinoma of the oral tongue with prognosis. International Journal of Surgical Pathology, 15(4), 376–383. Ebrahimi, A., Gil, Z., Amit, M., Yen, T.-C., Liao, C.-T., Chaturvedi, P., & Cernea, C. R. (2014). Primary tumor staging for oral cancer and a proposed modification incorporating depth of invasion: An international multicenter retrospective study. JAMA Otolaryngology-Head & Neck Surgery, 140(12), 1138–1148. Edge, S. B., & Compton, C. C. (2010). The American Joint Committee on Cancer: The 7th edition of the AJCC cancer staging manual and the future of TNM. Annals of Surgical Oncology, 17(6), 1471–1474. Effiom, O. A., Adeyemo, W. L., Omitola, O. G., Ajayi, O. F., Emmanuel, M. M., & Gbotolorun, O. M. (2008). Oral squamous cell carcinoma: a clinicopathologic review of 233 cases in Lagos, Nigeria. Journal of Oral and Maxillofacial Surgery, 66(8), 1595–1599. Ernst, J., Ikenberg, K., Apel, B., Schumann, D. M., Huber, G., Studer, G., & Korol, D. (2016). Expression of CK19 is an independent predictor of negative outcome for patients with squamous cell carcinoma of the tongue. Oncotarget, 7(46), 76151. Fillies, T., Jogschies, M., Kleinheinz, J., Brandt, B., Joos, U., & Buerger, H. (2007). Cytokeratin alteration in oral leukoplakia and oral squamous cell carcinoma. Oncology Reports, 18(3), 639–643. Fonseca, L.d. S., & Do Carmo, M. (2001). Identification of the AgNORs, PCNA and ck16 proteins in oral lichen planus lesions. Oral Diseases, 7(6), 344–348. Frohwitter, G., Buerger, H., Van Diest, P. J., Korsching, E., Kleinheinz, J., & Fillies, T. (2016). Cytokeratin and protein expression patterns in squamous cell carcinoma of the oral cavity provide evidence for two distinct pathogenetic pathways. Oncology Letters, 12(1), 107–113. Garcia, N. G., Oliveira, D. T., Lauris, J. R. P., Domingues, M. A. C., Minicucci, E. M., & Soares, C. T. (2016). Loss of cytokeratin 10 indicates malignant transformation in actinic cheilitis. Clinical Oral Investigations, 20(4), 745–752. Gonzalez‐Moles, M. A., Ruiz‐Avila, I., Rodriguez‐Archilla, A., & Martinez‐Lara, I. (2000). Suprabasal expression of Ki‐67 antigen as a marker for the presence and severity of oral epithelial dysplasia. Head & Neck: Journal for the Sciences and Specialties of the Head and Neck, 22(7), 658–661. Han, X., Zhao, J., Ji, Y., Xu, X., & Lou, W. (2013). Expression of CK19 and KIT in resectable pancreatic neuroendocrine tumors. Tumor Biology, 34(5), 2881–2889. Harris, T. M., Du, P., Kawachi, N., Belbin, T. J., Wang, Y., Schlecht, N. F., & Smith, R. V. (2014). Proteomic analysis of oral cavity squamous cell carcinoma specimens identifies patient outcome–associated proteins. Archives of Pathology & Laboratory Medicine, 139(4), 494–507. Jainkittivong, A., Swasdison, S., Thangpisityotin, M., & Langlais, R. P. (2009). Oral squamous cell carcinoma: A clinicopathological study of 342 Thai cases. The Journal of Contemporary Dental Practice, 10(5). Jiang, R., Gu, X., Moore-Medlin, T. N., Nathan, C.-A., & Hutt-Fletcher, L. M. (2012). Oral dysplasia and squamous cell carcinoma: Correlation between increased expression of

7


R.A. Safadi et al.

Systems Retrieved from https://seer.cancer.gov/canques/]. Nemes, J., Boda, R., Redl, P., & Marton, I. (2006). Oral squamous cell carcinoma in northeastern Hungary. II. Etiological factors. Fogorvosi szemle, 99(5), 179–185. Nyi, M. N., Gibson, T., Hanchard, B., Waugh, N., & McNaughton, D. (2014). Trends in incidence and age distribution of oral cavity and oropharyngeal squamous cell carcinomas, Kingston and St Andrew, Jamaica, 1978–2007. The West Indian Medical Journal, 63(2), 128. Pires, F. R., Ramos, A. B., Oliveira, J. B. C.d., Tavares, A. S., Luz, P. S. R.d., & Santos, T. C. R. B.d. (2013). Oral squamous cell carcinoma: Clinicopathological features from 346 cases from a single oral pathology service during an 8-year period. Journal of Applied Oral Science, 21(5), 460–467. Ruifrok, A. C., & Johnston, D. A. (2001). Quantification of histochemical staining by color deconvolution. Analytical and Quantitative Cytology and Histology, 23(4), 291–299. Safadi, R. A., Musleh, A. S., Al-Khateeb, T. H., & Hamasha, A. A.-H. (2010). Analysis of immunohistochemical expression of k19 in oral epithelial dysplasia and oral squamous cell carcinoma using color deconvolution-image analysis method. Head and Neck Pathology, 4(4), 282–289. Siriwardena, S. B., Tsunematsu, T., Qi, G., Ishimaru, N., & Kudo, Y. (2018). Invasionrelated factors as potential diagnostic and therapeutic targets in oral squamous cell carcinoma—A review. International Journal of Molecular Sciences, 19(5), 1462. Son, E.-M., Kim, J. Y., An, S., Song, K.-B., Kim, S. C., Yu, E., ... Hong, S.-M. (2015). Clinical and prognostic significances of cytokeratin 19 and KIT expression in surgically resectable pancreatic neuroendocrine tumors. Journal of Pathology and Translational Medicine, 49(1), 30. Stell, P. (1992). Prognosis in mouth cancer: Host factors. The Journal of Laryngology and Otology, 106(5), 399–402. van der Velden, L.-A., Manni, J., Ramaekers, F., & Kuijpers, W. (1999). Expression of intermediate filament proteins in benign lesions of the oral mucosa. European Archives of Oto-Rhino-Laryngology, 256(10), 514–519. Varsha, B. K., Radhika, M. B., Makarla, S., Kuriakose, M. A., Kiran, G. S., & Padmalatha, G. V. (2015). Perineural invasion in oral squamous cell carcinoma: Case series and review of literature. Journal of Oral and Maxillofacial Pathology: JOMFP, 19(3), 335–341. https://doi.org/10.4103/0973-029x.174630. Wang, S., Liu, B.-Y., & Chiang, C.-P. (2006). Decreased cytokeratin 13 expression in oral submucous fibrosis. Journal of Dentistry, 1(2), 59. Warnakulasuriya, S. (2009). Global epidemiology of oral and oropharyngeal cancer. Oral Oncology, 45(4-5), 309–316. Yoshida, K., Sato, K., Tonogi, M., Tanaka, Y., Yamane, G.-y., & Katakura, A. (2015). Expression of cytokeratin 14 and 19 in process of oral carcinogenesis. The Bulletin of Tokyo Dental College, 56(2), 105–111. Zhong, L.-p., Chen, W.-t., Zhang, C.-p., & Zhang, Z.-y. (2007). Increased CK19 expression correlated with pathologic differentiation grade and prognosis in oral squamous cell carcinoma patients. Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology, and Endodontology, 104(3), 377–384.

CD21, Epstein-Barr virus and CK19. Oral Oncology, 48(9), 836–841. Kale, A. D., Mane, D. R., Babji, D., & Gupta, K. (2012). Establishment of field change by expression of cytokeratins 8/18, 19, and MMP-9 in an apparently normal oral mucosa adjacent to squamous cell carcinoma: A immunohistochemical study. Journal of oral and maxillofacial pathology: JOMFP, 16(1), 10. Kaliszewski, K., Diakowska, D., Strutynska-Karpinska, M., Rzeszutko, M., Grzegrzolka, J., Dziegiel, P., & Sutkowski, K. (2016). Expression of cytokeratin-19 (CK19) in the classical subtype of papillary thyroid carcinoma: The experience of one center in the Silesian region. Folia histochemica et cytobiologica, 54(4), 193–201. Krishnan, L., Karpagaselvi, K., Kumarswamy, J., Sudheendra, U., Santosh, K., & Patil, A. (2016). Inter-and intra-observer variability in three grading systems for oral epithelial dysplasia. Journal of Oral and Maxillofacial Pathology: JOMFP, 20(2), 261. Laronde, D. M., Hislop, T. G., Elwood, J. M., & Rosin, M. P. (2008). Oral cancer: Just the facts. Journal of the Canadian Dental Association, 74(3). Le Campion, A. C. O. V., Ribeiro, C. M. B., Luiz, R. R., da Silva, F. F., Júnior, Barros, H. C. S., dos Santos, K.d. C. B, & Ferreira, S. M. S. (2017). Low survival rates of oral and oropharyngeal squamous cell carcinoma. International Journal of Dentistry, 2017. Lee, C.-C., Ho, H.-C., Su, Y.-C., Lee, M.-S., Hung, S.-K., & Chen, Y.-L. (2015). The prognostic ability of log odds of positive lymph nodes in oral cavity squamous cell carcinoma. Medicine, 94(27). Lee, Y.-M., Song, D. E., Kim, T. Y., Sung, T.-Y., Yoon, J. H., Chung, K.-W., ... Hong, S. J. (2016). Risk factors for distant metastasis in patients with minimally invasive follicular thyroid carcinoma. PLoS One, 11(5), e0155489. Liu, F., Chen, F., Huang, J., Yan, L., Liu, F., Wu, J., & Lin, L. (2017). Prospective study on factors affecting the prognosis of oral cancer in a Chinese population. Oncotarget, 8(3), 4352. Lydiatt, W. M., Patel, S. G., O’Sullivan, B., Brandwein, M. S., Ridge, J. A., Migliacci, J. C., ... Shah, J. P. (2017). Head and neck cancers—Major changes in the American Joint Committee on cancer eighth edition cancer staging manual. CA: A Cancer Journal for Clinicians, 67(2), 122–137. Makino, T., Yamasaki, M., Takeno, A., Shirakawa, M., Miyata, H., Takiguchi, S., & Matsuura, N. (2009). Cytokeratins 18 and 8 are poor prognostic markers in patients with squamous cell carcinoma of the oesophagus. British Journal of Cancer, 101(8), 1298. Maleki, D., Ghojazadeh, M., Mahmoudi, S.-S., Mahmoudi, S.-M., Pournaghi-Azar, F., Torab, A., & Naghavi-Behzad, M. (2015). Epidemiology of oral cancer in Iran: A systematic review. Asian Pacific Journal of Cancer Prevention, 16(13), 5427–5432. Moore, S., Pierce, A., & Wilson, D. (2000). ‘Oral cancer’—The terminology dilemma. Oral Diseases, 6(3), 191–193. Nanda, K. D. S., Ranganathan, K., Devi, U., & Joshua, E. (2012). Increased expression of CK8 and CK18 in leukoplakia, oral submucous fibrosis, and oral squamous cell carcinoma: an immunohistochemistry study. Oral Surgery, Oral Medicine, Oral Pathology and Oral Radiology, 113(2), 245–253. National Cancer Institute, S., Epidemiology and End Results Program. Cancer Query

8