Indian Journal of Pathology and Microbiology

ORIGINAL ARTICLE
Year
: 2020  |  Volume : 63  |  Issue : 3  |  Page : 369--375

Evaluation of the presence of myofibroblasts and matrix metalloproteinase 1 expression in the stroma of oral verrucous hyperplasia and verrucous carcinoma


Nazanin Mahdavi1, Pouyan Aminishakib1, Parisa Nabiyi2, Alireza Ghanadan3, Maedeh Ghorbanpour4, Merva Soluk-Tekkesin5,  
1 Department of Oral and Maxillofacial Pathology, School of Dentistry, Tehran University of Medical Sciences, Tehran, Iran
2 Graduated Student, School of Dentistry, Tehran University of Medical Sciences, Tehran, Iran
3 Department of Dermatopathology, Razi Skin Hospital, Tehran, Iran
4 Department of Oral and Maxillofacial Pathology, School of Dentistry, Islamic Azad University, Tehran, Iran
5 Institute of Oncology, Department of Tumor Pathology, Istanbul University, Istanbul, Turkey

Correspondence Address:
Merva Soluk-Tekkesin
Institute of Oncology, Department of Tumor Pathology, Istanbul University, Floor 2, Capa, 34093, Istanbul
Turkey

Abstract

Background: Oral verrucous carcinoma is a low-grade subtype of oral squamous cell carcinoma that should be differentiated from oral verrucous hyperplasia, a premalignant lesion. Stromal activated myofibroblasts known as cancer-associated fibroblasts have an active role in the initiation and progression of the cancers via secretion of different molecules including matrix metalloproteinases. Aims: This study is designed to understand the differences in the presence of myofibroblasts and expression of matrix metalloproteinase-1 in the adjacent stroma of verrucous carcinoma and oral verrucous hyperplasia (OVH). Settings and Design: Cross-sectional study. Material and Methods: Twenty-seven OVH, 19 oral verrucous carcinoma (OVC), and 8 cutaneous verrucous carcinoma (CVC) specimens were analyzed for immunohistochemical (IHC) expression of α-smooth muscle actin (αSMA) and MMP-1. Results: IHC studies for αSMA expression in nonvascular stromal cells of the adjacent stroma revealed mild or no expression in 81.4%, 73.7%, and 62.5% of the cases of OVH, OVC, and CVC groups, respectively. No significant difference was seen in αSMA expression index between OVH and OVC groups (Adj. Sig. = 0.220) and between OVC and CVC groups (Adj. Sig. = 1.00). Pairwise analysis revealed a significant difference in MMP-1 expression index between the groups. No significant correlation was observed between MMP-1 expression index and αSMA expression index in OVH (pv = 0.358) and OVC (pv = 0.388) groups. Conclusion: The differences in MMP-1 expression between OVH and OVC can be used as an adjunctive aid in challenging cases including disoriented or inadequate samples.



How to cite this article:
Mahdavi N, Aminishakib P, Nabiyi P, Ghanadan A, Ghorbanpour M, Soluk-Tekkesin M. Evaluation of the presence of myofibroblasts and matrix metalloproteinase 1 expression in the stroma of oral verrucous hyperplasia and verrucous carcinoma.Indian J Pathol Microbiol 2020;63:369-375


How to cite this URL:
Mahdavi N, Aminishakib P, Nabiyi P, Ghanadan A, Ghorbanpour M, Soluk-Tekkesin M. Evaluation of the presence of myofibroblasts and matrix metalloproteinase 1 expression in the stroma of oral verrucous hyperplasia and verrucous carcinoma. Indian J Pathol Microbiol [serial online] 2020 [cited 2020 Oct 29 ];63:369-375
Available from: https://www.ijpmonline.org/text.asp?2020/63/3/369/291675


Full Text



 Introduction



Oral verrucous carcinoma (OVC) is a low-grade subtype of oral squamous cell carcinoma (OSCC) with very limited capacity of invasion and rare metastatic potential.[1] Despite being classified as a carcinoma, this tumor shows overlapping features with benign verruci-form lesions.[2] OVC should be clinically differentiated from proliferative verrucous leukoplakia, leukoplakia, oral verrucous hyperplasia (OVH), and oral squamous cell carcinoma (OSCC).[3]

One of the most important differential diagnoses of OVC is OVH, which shares many similar clinical and histological features with this tumor. OVH is a potentially premalignant lesion of the oral cavity with a malignant transformation rate of 3% to 17%.[4]

Both OVH and OVC clinically present as an extensive thick white plaque or mass with exophytic verrucous or papillary surface projections. Thus, there are no striking clinical characteristic features to distinguish OVH from OVC. By now evaluation of the hematoxylin and eosin (H and E) staining slides is the most reliable method to distinguish the two lesions.[5],[6]

Histopathologically, both lesions show marked acanthosis with papillary surface projections and a highly keratinized surface. Bulbous rete ridges, which are deeply extended to the underlying connective tissue, demonstrating pushing borders (rather than infiltrating into the underlying connective tissue) with an intact basement membrane are important features of VC. Lack of significant atypia in the epithelial cells is an important feature of OVC.[1],[7]

Unlike OVC, OVH does not show downward growth into the connective tissue. The epithelial cells may have different levels of cellular atypia. Epithelial dysplasia is a prominent feature in 77.7% of OVH cases. Advanced lesions of OVH may progress to OSCC or OVC in a lower occurrence.[4],[7],[8],[9]

Although conventional therapy for either OVH or VC is a complete surgical excision, the fact that 20% of VC cases have foci of SCC and need for more aggressive treatment emphasizes on the importance of distinguishing these two entities.[10] Nonetheless, histopathological differentiation between these two lesions still is challenging, especially for the case when the biopsy sample has improper orientation or does not contain neighboring normal epithelium. In order to differentiate OVC from OVH, epithelial features do not clearly resolve the issue, if the lesions lack malignant cytology.[10]

During the progression of cancer, tumor cells are able to alter the characteristics of the adjacent stroma to create a supportive microenvironment.[11],[12] The creation and accumulation of a special type of reactive fibroblasts in the stroma is an early phenomenon in the Carcinoma in situ. These reactive fibroblasts, which share a similar morphology with myofibroblasts observed in tumors, are called cancer-associated fibroblasts (CAFs). The results of several in vitro and in vivo experiments indicate that CAFs promote cancer progression in both proliferation and invasion through multiple growth factors and signaling pathways.[10],[11],[12]

Although CAFs are identified using several markers such as tenascin-C, periostin, neuron glial antigen-2 (NG2), vimentin, desmin, platelet-derived growth factor receptor-α and β (PDGFR α and β), and fibroblast specific protein-1 (FSP-1), they are not necessarily specific for myofibroblasts. In addition, fibroblast activation protein (FAP) is another useful marker for CAFs and myofibroblasts but this marker is also detected in malignant epithelial cells.[12],[13],[14] αSMA positivity is considered as a diagnostic factor for fibroblast activation during tumorigenesis. αSMA is the most widely used marker for CAFs. It has been known as a specific marker for myofibroblasts, which are considered to be activated fibroblasts.[14] Based on studies, oral lesions are rich in fibroblastic stroma and the main αSMA positive cells are the pericytes. Therefore, αSMA-positive activated myofibroblasts can easily be detectable.[13]

CAFs are also responsible for secreting enzymes of matrix remodeling, including the matrix metalloproteinases (MMPs), a family of 23 human zinc dependent proteolytic enzymes that specifically degrade the extracellular matrix (ECM) and basement membrane (BM) during tumor cell invasion. MMP-1 which belongs to the collagenase group of this family breaks down the first, second, third, and tenth collagen types. First and third types are found extensively in connective tissue.[15],[16]

By now researches have mainly focused on the epithelial characteristics of OVC and OVH. Therefore, this study is designed to investigate the differences in the presence of activated fibroblasts and expression of MMP-1 in the adjacent connective tissue of verrucous carcinoma and verrucous hyperplasia. To the best of our knowledge, this is the first study that evaluates and compares the expression of αSMA and MMP-1 in the stroma of OVH and VC.

 Material and Methods



Samples selection

The protocol for this study was approved by the University Medical Sciences Research and Ethics Committee [IR.TUMS.VCR.REC.1395.537]. This study performed on paraffin-embedded blocks with the diagnosis of OVH, OVC, and cutaneous verrucous carcinoma (CVC) from the pathology archives of the hospital. Two expert oral pathologists independently and blindly re-examined H and E stained slides to ensure conformity with accepted diagnostic criteria. The histologic diagnostic criterion used in our study for OVH was Shear and Pindborg's description for both sharp and blunt forms of OVH.[4],[7],[9],[17] For the diagnosis of OVC, the following criteria were used: a well-differentiated proliferation of squamous epithelium with a highly keratinized surface, lack of significant atypia in the epithelial cells or a mild pleomorphism in the basal and supra-basal cell layers, broad bulbous rete ridges which are deeply extended to the underlying connective tissue or pushing borders, and a mixed chronic inflammatory cell infiltrate which may also be prominent in the stroma.[2],[7],[17]

27 cases of OVH and 27 cases of VC (19 OVCs and 8 CVCs) between 1999 and 2017 collected. For the shortage of our sample number in VC group, CVC specimens were decided to include the study due to the same nature with different localizations that there are not any published data about the different nature of OVC and CVC yet. For each lesion, demographic information, clinical and histopathological data, and IHC results were recorded in a separate form. The degree of epithelial dysplasia was noted as carcinoma in-situ for the presence of dysplastic cells in the full thickness of the epithelium, severe dysplasia for the presence of dysplastic cells beyond the middle of barbed layer (stratum spinosum), moderate dysplasia for the presence of dysplastic cells up to the middle of the barbed layer, mild dysplasia for the presence of dysplastic cells in the basal and para-basal layers, and without dysplasia for lack of dysplastic features.[7]

Immunohistochemistry

Formalin-fixed paraffin-embedded blocks were cut into 4 μm sections and stained immunohistochemically for αsma (BIOCARE, 1A4, Mouse Monoclonal, Concentrated and Prediluted, Ready to Use, PH = 6.0) and MMP-1 (GeneTex, Rabbit Polyclonal, GTX100534, 1:200, PH = 9.0) markers, manually according to the supplier's recommendation.

Colon adenocarcinoma and leiomyoma of the uterus were used as positive control for MMP-1 and αSMA, respectively, and wall of the well-formed blood vessels served as positive internal controls for αSMA. For negative, control sections were treated with normal saline and were confirmed to be unstained.

Scoring

For determining MMP-1 expression level, an investigator, who was not informed about the clinical data and type of the specimen, calculated the number of positive stromal cells per 100 stromal cells (0%–100%) from three separate cell-rich areas at ×400 magnification and the mean value of positive areas was recorded.[18]

The intensity of αSMA and MMP-1 immunoreactivities semi-quantitatively was evaluated using the following intensity categories: 0 (no staining), 1+ (weak but detectable staining), 2+ (moderate or distinct staining), and 3+ (intense staining).

For determining αSMA expression level, cytoplasmic staining of the stromal cells regardless of the intensity of the staining was considered positive and the percentage of the stained cells recorded and graded on a scale of 0–3: 0 = 0% positive cells, 1 = 1%–33% positive cells, 2 = 34%–66% positive cells, and 3 = 67%–100% positive cells.[13]

For each sample, the staining percentage and staining intensity scores were multiplied to give the staining index; 0 = zero, 1–2 = low, 3–4 = moderate, 6–9 = high.

For the assessment of the relation of combination of the MMP-1 and αSMA markers, we scored MMP-1 expression as follows: Point 0: ≤ 5%, 6% ≤ 30%, 31% point 2 ≤70%, and 71% ≤ 100%.[19]

Statistical analyses

Mann–Whitney U test was used to examine OVH and VC groups with the studied variables. To compare variables between the three groups of the lesions (OVH, OVC, and CVC), Kruskal–Wallis test was used. When the comparison between the three groups (OVH, OVC, and CVC) became significant, pair-wise comparisons were performed using the Bonferroni correction method. Spearman's correlation test was used to investigate the relationship between ordinal and quantitative variables. The level of statistical significance is at P value (PV) <0.05.

 Results



This study was performed on 27 cases of OVH and 27 cases of VC (19 cases of OVC and 8 cases of CVC) [Figure 1]a, [Figure 2]a and [Figure 3]a.{Figure 1}{Figure 2}{Figure 3}

IHC studies for αSMA expression in nonvascular stromal cells of the adjacent stroma revealed that 81.4% of the OVH cases showed low or no αSMA expression [Figure 1]b. In the OVC and CVC groups, 73.7% and 62.5% of the cases had mild or no expression, respectively [Table 1] and [Figure 2]b, [Figure 3]b. No significant difference was seen in αSMA expression index between OVH and OVC groups (Adj. Sig. = 0.220) and between OVC and CVC groups (Adj. Sig. = 1.00) [Table 2].{Table 1}{Table 2}

Regarding the immunohistochemical expression of MMP-1, none of OVH cases revealed moderate or high MMP-1 expression index while 15.78% and 37.5% of OVC and CVC cases had moderate or high expression index [Table 1] and [Figure 1]a, [Figure 1]b, [Figure 1]c. Pairwise analysis revealed a significant difference in MMP-1 expression index between the groups [Table 2].

No significant correlation was observed between MMP-1 expression index and αSMA expression index in OVH (pv = 0.358) and OVC (pv = 0.388) groups [Table 3].{Table 3}

This study was designed to evaluate MMP-1 expression in the stromal fibroblasts, but during the evaluation of IHC slides, we found positive MMP-1 expression in the basal cell layer of the epithelium in some cases and recorded it as positive and negative [Figure 1]c. The mean percentage of MMP-1 positivity in the basal cells was 40.7% for OVH and 21.1% and 37.5% for OVC and CVC cases, respectively [Table 1].

 Discussion



OVC is a low-grade subtype of OSCC, which was first described by Ackerman in 1948.[1] One of the most important differential diagnoses of OVC is OVH which is a potentially premalignant lesion of the oral cavity with a malignant transformation rate of 3% to 17% and shares many similar clinical and histological features with OVC.[4],[20],[21] Some researchers believe that OVH and VC are histologically different lesions.[22] Others considered OVH as a spectrum of VC and did not agree to respect OVH as a separate lesion.[23] In addition, Murrah and Batsakis consider OVH as an irreversible lesion, which is precedent of VC.[24]

This study is performed on 27 cases of OVH and 27 cases of VC (19 cases of OVC and 8 cases of CVC). Histopathologically, in our study, epithelial dysplastic features were significantly higher in OVHs than VC group. In line with these findings, the previous studies reported higher prevalence of epithelial dysplasia in OVH.[25]

A number of studies confirmed the presence of a specific type of stromal fibroblast with myoepithelial feature called CAFs in the carcinomatous stroma and their important role in the development and progression of malignant tumors.[11],[12],[13] Activated stromal myofibroblastic cells that might be pioneer of CAFs were immunohistochemically identified by the presence of αSMA in our research. According to the results, about 92.6% of VCs and 59.2% of OVHs showed αSMA positive cells in their stroma. Chatzistamou et al. and Chaudhary et al. demonstrated significantly higher αSMA expression in the stroma of VCs as a low-grade subtype of a malignant lesion, compared to OVH as a potentially premalignant lesion. Paral et al. reported high number of αSMA positive cells in the stroma of OVC and no expression in OVH.[10],[13],[26] Rodrigues et al. failed to demonstrate the presence of stromal myofibroblasts in verrucous carcinoma.[27]

We found that 81.4% of OVH cases showed no or mild αSMA expression index. Different studies have emphasized the role of CAFs in the development and progression of different cancer types including OSCCs.[10],[12],[13],[14],[26] Our findings are aligned with the results of Rodrigues et al. and indicate that CAFs-like activated myofibroblastic cells have limited role in the development of OVH and further studies are needed to assess the role of these cells in the progression of SCC after the initiation of the tumor.

The ability of tumor cells to infiltrate surrounding tissue is an important feature of malignant potential. This process is based on the ability of the tumor to degrade BM and ECM including the BM. Many previous studies have identified MMPs, a family of zinc-dependent endopeptidases, to have a key role in this process. The proteolytic mechanism of MMPs seems to be crucial for tumor invasion and its metastatic expansion.[16]

Structural changes in ECM by MMPs are an important step in tumor progression. The role of CAFs in tumor invasion has been established and it seems that this process is regulated by activating MMPs-degrading ECM and followed by angiogenesis and metastasis of tumor cells. Therefore, MMPs expression pattern clearly correlates with the histopathological features and clinical behavior of oral carcinoma and the activated fibroblasts conduct invasion of the cancer cells by producing tunnels through ECM.[12],[28] Some studies emphasized on the role of CAFs in the progression of oral SCCs and found higher number of CAFs in the stroma of OSCCs than VCs and other precancerous lesions.[12],[28]

Histopathologically, OVC demonstrates pushing borders with an intact BM, which is thicker in OVC than OSCC. Unlike OSCC, VC does not infiltrate in the underlying connective tissue by forming tumor islands and nests and its characteristic pushing border has been attributed to the presence of an intact basement membrane.[7] Arduino et al. indicate dissimilar expression of some proteins of the BM in various conditions. The staining pattern of BM in the VC was more often continuous than its pattern in the preneoplastic lesion and was always well expressed, which might explain the low malignant potential and poor ability to invade surrounding tissues.[15] The results of Kadeh et al. study demonstrate that the lower growth nature of OVC invasion, as compared with that of OSCC, may be related to its MMPs expression pattern.[29] Impola et al. suggested that a difference in MMP expression between OSCC and VC is responsible for the differences in their biological behavior. They found out the ability of epithelial cells for degradation of the basement membrane and invading the connective tissue is related to secretion of specific types of MMPs including MMP-7, -9, and 12 and this MMP expression profile is not seen in VCs.[30]

We found a significantly higher MMP-1 expression in the stroma of VC compared to OVH which can indicate the different nature of these lesions. Klieb et al. reported similar findings.[17] A number of the previous studies reported MMP-1 overexpression in stromal fibroblasts of OSCC. A possible explanation for this difference is that MMP-1 does not have a significant role in the destruction of the basement membrane and initial invasion of the tumor; however, after the tumor invasion, MMP-1 may have an active role in the degradation of ECM and tumor progression as expression of MMP-1 and its association with the poor prognosis has been shown in a number of tumors. In contrast to these findings, Jordan et al. found that the elevated levels of MMP-1 mRNA are associated with oral dysplasias that progress into oral cancer compared with those dysplasias did not progress into malignancy.[16],[27]

Although oral verrucous carcinoma and cutaneous verrucous carcinoma are the same entities with different involved locations, the results of the present study demonstrated some other features that are not the same across these two lesions. We observed that the minimum age at the diagnosis of OVC is 32 years and for CVC the minimum age is 50 years. In addition, a marked male predominance is found in CVCs compared to OVCs. Dysplastic features of the epithelium are greater in CVCs compared to OVCs. Furthermore, these lesions demonstrated different MMP-1 expression levels. Therefore, the likelihood that these two lesions might be different phenomena can be taken into consideration and require further investigations.

Focus of the research was on OVH and OVC according to the challenging histopathological differentiation between these two lesions, especially for the case when the biopsy sample has improper orientation or does not contain neighboring normal epithelium and also the fact that as many as 20% of VC lesions have a conventional SCC developing concurrently within a VC and the need for immediate and the more aggressive treatment.[4] To the best of our knowledge, this is the first study that evaluates and compares the simultaneous expression of αSMA and MMP-1 in the stroma of OVH and VC. In this study, 40.7% of OVC cases had zero scores for both MMP-1 and αSMA and among the intraoral lesions (OVC and OVH), 94.7% of the cases who reviled negative scores for both markers were OVH, which indicates that using negative expression of both markers has low sensitivity and high specificity for the diagnosis of OVC from OVH and can be used as an adjunctive aid in challenging cases including disoriented or inadequate samples.

 Conclusion



In conclusion, the results herein show that the expression of MMP-1 in the stroma of VC is significantly higher than OVH and there is no significant correlation between the presence of myofibroblasts and the expression of MMP-1 in these lesions. Further studies are recommended to investigate the expression of the different subtypes of MMPs and their relationship with myofibroblasts in OSCC, OVH, and VC to clarify the role of MMP profile and myofibroblasts in the initiation and progression of oral cancers.

Financial support and sponsorship

This study was supported by a grant from Tehran University of Medical Sciences (grant number: 32272).

Conflicts of interest

There are no conflicts of interest.

References

1Rekha K, Angadi PV. Verrucous carcinoma of the oral cavity: A clinico-pathologic appraisal of 133 cases in Indians. Oral Maxillofac Surg 2010;14:211-8.
2Shergill AK, Solomon MC, Carnelio S, Kamath AT, Aramanadka C, Shergill GS. Verrucous carcinoma of the oral cavity: Current concepts. Int J Sci Stud 2015;3:114-8.
3Mehrotra D, Goel M, Kumar S, Pandey R, Ram H. Oral verrucous lesions: Controversies in diagnosis and management. J Oral Biol Craniofac Res 2012;2:163-9.
4Sharma P, Wadhwan V, Aggarwal P, Sharma A. Oral verrucous hyperplasia versus oral verrucous carcinoma: A clinicopathologic dilemma revisited using p53 as immunohistochemical marker. J Oral Maxillofac Pathol 2016;20:362.
5Kallarakkal TG, Ramanathan A, Zain RB. Verrucous papillary lesions: Dilemmas in diagnosis and terminology. Int J Dent 2013;2013. doi: 10.1155/2013/298249.
6Zhu LK, Ding YW, Liu W, Zhou YM, Shi LJ, Zhou ZT. A clinicopathological study on verrucous hyperplasia and verrucous carcinoma of the oral mucosa. J Oral Pathol Med 2012;41:131-5.
7Neville BW, Damm DD, Chi AC, Allen CM. Oral and Maxillofacial Pathology. Elsevier Health Sciences; 2015.
8El-Rouby DH. Association of macrophages with angiogenesis in oral verrucous and squamous cell carcinomas. J Oral Pathol Med 2010;39:559-64.
9Navaneetham A, Saraswathi MD, Santosh B. Oral verrucous hyperplasia: A case report. J Maxillofac Oral Surg 2014;13:346-8.
10Paral KM, Taxy JB, Lingen MW. CD34 and α smooth muscle actin distinguish verrucous hyperplasia from verrucous carcinoma. Oral Surg, Oral Med, Oral Pathol Oral Radiol 2014;117:477-82.
11San Martin R, Barron DA, Tuxhorn JA, Ressler SJ, Hayward SW, Shen X, et al. Recruitment of CD34+ fibroblasts in tumor-associated reactive stroma: The reactive microvasculature hypothesis. Am J Pathol 2014;184:1860-70.
12Xing F, Saidou J, Watabe K. Cancer associated fibroblasts (CAFs) in tumor microenvironment. Front Biosci (Landmark Ed) 2010;15:166-69.
13Chaudhary M, Gadbail AR, Vidhale G, Mankar MP, Gondivkar SM, Gawande M, et al. Comparison of myofibroblasts expression in oral squamous cell carcinoma, verrucous carcinoma, high risk epithelial dysplasia, low risk epithelial dysplasia and normal oral mucosa. Head Neck Pathol 2012;6:305-13.
14Shiga K, Hara M, Nagasaki T, Sato T, Takahashi H, Takeyama H. Cancer-associated fibroblasts: Their characteristics and their roles in tumor growth. Cancers 2015;7:2443-58.
15Arduino PG, Carrozzo M, Pagano M, Broccoletti R, Scully C, Gandolfo S. Immunohistochemical expression of basement membrane proteins of verrucous carcinoma of the oral mucosa. Clin Oral Invest 2010;14:297-302.
16Jordan RC, Macabeo-Ong M, Shiboski CH, Dekker N, Ginzinger DG, Wong DT, et al. Overexpression of matrix metalloproteinase-1 and-9 mRNA is associated with progression of oral dysplasia to cancer. Clin Cancer Res 2004;10:6460-5.
17Klieb HBE, Raphael SJ. Comparative study of the expression of p53, Ki67, E-cadherin and MMP-1 in verrucous hyperplasia and verrucous carcinoma of the oral cavity. Head Neck Pathol 2007;1:118-22.
18Boström P, Söderström M, Vahlberg T, Söderström KO, Roberts PJ, Carpén O, et al. MMP-1 expression has an independent prognostic value in breast cancer. BMC Cancer 2011;11:348.
19Wang YD, Yan PY. Expression of matrix metalloproteinase-1 and tissue inhibitor of metalloproteinase-1 in ulcerative colitis. World J Gastroenterol 2006;12:6050-3.
20Hazarey VK, Ganvir SM, Bodhade AS. Verrucous hyperplasia: A clinico-pathological study. J Oral Maxillofac Pathol 2011;15:187.
21Wang YP, Chen HM, Kuo RC, Yu CH, Sun A, Liu BY, et al. Oral verrucous hyperplasia: Histologic classification, prognosis, and clinical implications. J Oral Pathol Med 2009;38:651-6.
22Arendorf T, Aldred M. Verrucous carcinoma and verrucous hyperplasia. J Dent Assoc S Afr 1982;37:529-32.
23Slootweg PJ, Müller H. Verrucous hyperplasia or verrucous carcinoma: An analysis of 27 patients. J Maxillofac Surg 1983;11:13-9.
24Murrah VA, Batsakis JG. Proliferative verrucous leukoplakia and verrucous hyperplasia. Ann Otol, Rhinol Laryngol 1994;103:660-3.
25Hosseinpour S, Mashhadiabbas F, Ahsaie MG. Diagnostic biomarkers in oral verrucous carcinoma: A systematic review. Pathol Oncol Res 2017;23:19-32.
26Chatzistamou I, Dioufa N, Trimis G, Sklavounou A, Kittas C, Kiaris H, et al. p21/waf1 and smooth-muscle actin α expression in stromal fibroblasts of oral cancers. Anal Cell Pathol 2010;33:19-26.
27Rodrigues PC, Da Costa Miguel MC, De Aquino SN, Fonseca FP, Dos Santos Silva AR, Paes Leme AF, et al. Stromal myofibroblasts in potentially malignant and malignant lesions of the oral cavity. Oncol Lett 2015;9:667-70.
28Brentnall TA. Arousal of cancer-associated stromal fibroblasts: Palladin-activated fibroblasts promote tumor invasion. Cell Adh Migr 2012;6:488-94.
29Kadeh H, Saravani S, Heydari F, Keikha M, Rigi V. Expression of matrix metalloproteinase-10 at invasive front of squamous cell carcinoma and verrucous carcinoma in the oral cavity. Asian Pac J Cancer Prev 2015;16:6609-13.
30Impola U, Uitto VJ, Hietanen J, Hakkinen L, Zhang L, Larjava H, et al. Differential expression of matrilysin-1 (MMP-7), 92 kD gelatinase (MMP-9), and metalloelastase (MMP-12) in oral verrucous and squamous cell cancer. J Pathol 2004;202:14-22.