Background

Squamous cell carcinoma (SCC) of the oral cavity is a severe health problem in terms of morbidity and mortality. It is the sixth most common cancer worldwide and accounts for 4% of all malignancies and 2% of malignancies in women [1, 2]. These tumors primarily develop in individuals with a protracted history of tobacco and alcohol abuse. However, the specific cause of the disease remains unknown [3].

Despite the evolving model of multimodality management, which integrates surgical intervention, chemotherapy, and radiation therapy, overall survival remains poor with a five-year relative survival rate below 50% [4].

The majority of oral SCC patients presents with advanced disease and incur significant morbidity and mortality due to the limited availability of screening tools and markers for adjuvant therapy. Progress has been made regarding the identification and validation of biomarkers as targeted molecular therapies have advanced through clinical trials. To determine which patients will benefit from these treatments, personalized medical methods and analysis of factors specific to the individual and their tumor are required [5].

The phosphatidylinositol 3 kinase (PI3K)/protein kinase B (AKT) signaling pathway plays important roles in regulating tumor cell survival, apoptosis, and protein translation [6, 7]. PI3K is activated by receptor tyrosine kinases, leading to allosteric activation of the enzyme and tyrosine phosphorylation of its regulatory subunit. In response to extracellular stimuli, PI3K phosphorylates the 3′-hydroxyl group of phosphatidylinositol-4, 5-bi-phosphate (PIP2) to generate phosphatidylinositol-3, 4, 5-triphosphate (PIP3). AKT is activated by phosphorylation via PIP3 and mammalian target of rapamycin (mTOR) and is a component of the mTOR complex [8]. The phosphatase and tensin homolog deleted on chromosome 10 (PTEN) is a well-described negative regulator of the PI3K/AKT signaling pathway, which functions as a tumor suppressor gene by decreasing the levels of cyclin D1 and inducing G1 phase cell cycle arrest [9].

The PI3K/AKT pathway is frequently overactive in various tumors, which triggers a cascade of responses that drive tumor progression, including aberrant cell growth, proliferation, increased cell survival, and increased cell motility [10]. Although a few reports have examined AKT, PTEN, and mTOR expression in head and neck SCC [11,12,13], there is no comprehensive study of these markers in rare groups of oral SCC that include gingival, palate hard, and alveolar ridge tumors.

In the present study, we have assessed the expression of PI3K/AKT pathway proteins in 125 oral SCCs using immunohistochemistry. Additionally, we examined the possible correlations between protein expression and clinicopathological parameters in this cohort of patients.

Methods

Patients

A total of 125 patients with oral SCC treated over a period from 1980 to 2006 were identified from files obtained from the Department of Pathology, A. C. Camargo Cancer Center (Sao Paulo, Brazil). The mean age of the patients was 65 years (range: 30–91 years), and samples were collected from 76 males and 49 females. The clinicopathological parameters available included age, gender, tumor location, clinical stage, and histological grade. Tissue slides were reviewed for histological classification according to WHO [14]. The series included 32 gingival, 31 hard palate, and 62 alveolar ridge tumors. Patient clinical outcome was followed from the date of diagnosis to a period of 0.6–108.6 months (mean of 28.3 months). Patients who were lost to follow-up or died from any cause other than oral SCC were excluded from the analysis of survival rates. No tumors associated with specific genetic mutations were included. The institutional board for ethical studies at the A. C. Camargo Cancer Center approved this study. All the clinical characteristics and pathological findings are summarized in Table 1.

Table 1 Clinicopathological characteristics of the 125 oral carcinomas patients
Full size table

Tissue microarray (TMA) construction

For TMA construction, paraffin blocks of formalin-fixed oral SCC were retrieved from the pathology archives. A slide containing a representative portion of the tumor was selected and an area of the tumor was circled on the slide. With the use of a tissue microarrayer (Beecher Instruments, Silver Spring, MD, EUA), the area of interest in the donor block was cored twice with a 1.0 mm diameter needle and was transferred to a recipient paraffin block. Cores were arrayed onto three separate blocks that contained two tumor cores from each case. An adequate case was defined as a tumor occupying >90% of the core area.

Immunohistochemistry

TMA sections were placed on adhesive coated slides for immunohistochemistry (IHC). A standard peroxidase-conjugated streptavidin-biotin method was used to detect the staining reaction (LSAB+, DAKO, Carpinteria, USA) according to the manufacturer’s protocol. The antibodies, dilutions, sources and positive controls are detailed in Table 2. Four-micrometer thick sections were cut from the array block, deparaffinized with xylene, and dehydrated through a series of graded alcohols. Microwave antigen retrieval was used for all antibodies and was performed by placing the slides in 10 mM citrate buffer (pH 6.0) for 15 min. For negative controls, the primary antibodies were omitted and PBS was used as a substitute.

Table 2 Primary antibodies, clones, dilutions, sources, and positive controls used in immunohistochemistry
Full size table

Evaluation of stained sections

The staining was evaluated by light microscopic examination and was interpreted by two independent pathologists who were blinded to the available clinical information. The proteins were observed in the nucleus, membrane, and cytoplasm of tumor cells. Protein analysis was performed for 10 randomly selected fields viewed at high-power. A total of 100 tumor cells were counted from each field for each case and for all antibodies analyzed. The percentage of positive cells was calculated by dividing the number of positive tumor cells by the total of tumor cells counted. Staining was considered to be positive when >10% of the cells were positive, this cut off was based in previous studies [15, 16].

Statistical analysis

The survival curves were estimated using the Kaplan-Meier product-limit method, and the significant differences between the survival curves were determined using the log-rank test. The multivariate survival analysis was performed using the Cox proportional hazards model. Correlation coefficients between immunohistochemical expression and clinicopathological findings were estimated by Pearson correlation. Either the chi-square test or Fisher’s exact test (two-sided) was performed to determine the correlation between protein expression and clinicopathological parameters. The results were considered statistically significant at p < 0.05. All statistical analyses were conducted using the SPSS 10.0 statistical software program (SPSS, Chicago, IL, USA).

Results

Expression of PI3K, p-AKT, p-mTOR, p-MAPK, and PTEN protein

The expression of p-AKT, p-mTOR, PTEN, PI3K, and Mitogen Activated Protein Kinases p-MAPK was observed in a considerable number of oral SCCs and these results are summarized in Table 3. For most of the antibodies tested, there were no statistically significant differences between immunostaining levels and tumor location in oral SCCs. Overall, 110 (92.4%) of the oral SCCs were positive for PI3K. We observed p-mTOR staining in the cytoplasm of the tumor cells in 105 (88.2%) cases, membrane staining in 53 cases (44.5%), and nuclear staining in 51 cases (42.9%). The majority of the cases (113 of 120) were positive for p-AKT in the cytoplasm of the tumor cells, and nuclear staining for this marker was observed in 77 (64.2%) of the cases. Cytoplasmic p-MAPK staining was found in 106 cases (88.3%), and 70 (58.3%) of the cases exhibited positive nuclear staining. PTEN expression was observed in the cytoplasm of the tumor cells in 30 cases (25%), and nuclear staining was observed in 35 (28%) cases. Figure 1 shows the positive oral SCC staining patterns for the markers studied by immunohistochemistry.

Table 3 Expression of the proteins studied by immunohistochemistry and association with tumor site of oral carcinomas
Full size table
Fig. 1
figure 1

Immunostaining of PI3K in oral SCC (a, original size 200×), pAKT overexpression (b, original 200×), p-MAPK (c, original size 200×), p-MTOR (d, original size 150×) and PTEN (e, original size 200×). The staining pattern is primarily in the membrane and/or cytoplasm of tumor cells (a, b, d), although the nuclear pattern can also be observed (c, e)

Full size image

Association between p-AKT, p-mTOR, PTEN, PI3K, and p-MAPK expression and clinicopathological factors

The expression levels of all proteins studied were not significantly associated with clinicopathological features, and only the p-MAPK positive staining was statistically associated with advanced tumors at clinical stages III and IV (P = 0.042).

The influence of overexpression of p-AKT, p-mTOR, PI3K, p-MAPK, and PTEN protein on survival in patients with oral SCC

Except for tumor histological grade (P = 0.001; Fig. 2), none of the clinicopathological features exhibited a significant influence on patient disease free survival analyses. In Cox’s multiple regression analysis, patients with nuclear p-mTOR overexpression tumors had a significantly better rate of disease free survival than patients without nuclear p-mTOR overexpression (P = 0.037; Fig. 3). The results of univariate analysis of disease-free survival rate associated to clinicopathological and molecular parameters are summarized on Table 4.

Fig. 2
figure 2

Kaplan –Meier curve for the disease free survival rates in relation to tumor histological grade. The difference between the patients with SCC II e III (green line) and patients with oral SCC I (blue line) is statiscally significant (long-rank test, p = 0.001)

Full size image
Fig. 3
figure 3

Kaplan –Meier curve for the disease free survival rates in relation to p-MTOR nuclear expression. The difference between the patients with p-MTOR positive (green line) and patients with p-MTOR negative oral SCC (blue line) is statiscally significant (long-rank test, p = 0.037)

Full size image
Table 4 Univariate analysis of disease-free survival rate according to clinicopathological findings and proteins expressions
Full size table

Discussion

The identification of prognostic and predictive markers is clinically important because oral cancer is a group of heterogeneous diseases with various biological and clinical characteristics.

In recent years, the PI3K/AKT/mTOR signaling pathway has been shown to be a key regulator of cell cycle proliferation, growth survival, protein synthesis, and glucose metabolism [2, 6, 17]. The significance of this pathway in cancer is underscored by abundant evidence suggesting that multiple components of this pathway are frequently deregulated and serve as molecular targets in cancer development and progression [18].

Exploring the regulation of the PTEN/PI3K pathway in cancer will lead to insight into the malignant behavior of tumor cells. Such information will be useful for the optimization of rationally designed cancer therapies [5, 19].

Few studies exist regarding the expression of the signaling proteins of the PI3K/AKT/mTOR pathway in head and neck SCC. Several mutations were identified in this pathway, including PI3K (8–10%), mTOR (5–8%), and PTEN (23%) [20,21,22]. AKT is known to be activated in the majority of head and neck SCC tumor tissue and cell lines and its active form is detected in 50% of preneoplastic lesions [23, 24].

Our study is the first report to show the expression of PI3K/AKT proteins in gingival, hard palate, and alveolar ridge SCCs. We observed a high frequency of positive staining for PI3K, p-AKT, p-mTOR, and p-MAPK in oral SCCs, including gingival, hard palate, and alveolar ridge SCCs. However, the majority of our cases showed loss of PTEN expression, suggesting that activation of PI3K/AKT pathway occurred downstream of PTEN. A possible role for PTEN loss in the pathogenesis of head and neck SCC has been described by Califano et al. in a stepwise model of carcinogenesis [25]. The majority of PTEN gene dysfunction has been attributed to mutation, loss of heterozygosity (LOH), and epigenetic silencing [24]. As we presumed in our cases, these disorders lead to a decrease or loss of protein expression, thus providing a rationale for defective PTEN detection by immunohistochemistry.

The prognosis in oral carcinomas varies according to the tumor location. Local and regional lymph node involvement is the most important prognostic factor. An increased risk is observed with larger tumors, whereas poorly differentiated tumors metastasize more frequently than well-differentiated tumors [3]. Similarly, the only clinicopathological factor associated with patient prognosis in our study was tumor grade. We found that poor prognosis was associated with poorly differentiated oral SCC. Some molecular indicators of prognosis have been observed in oral SCCs. The overexpression of Epidermal growth factor receptor (EGFR) and Transforming growth factor alpha (TGF-α) is associated with a decrease in the disease-free and cancer specific survival rates [26]. Cyclin D1 overexpression and lack of p16 expression correlate with a worse prognosis, increased recurrence, and poor survival rates in oral cancer [27]. Overexpression of bcl-2 and loss of PTEN expression were correlated with poorly differentiated tumors, lymph node involvement, and late stages in oral SCC [28]. We found that PTEN protein expression was lost in more than 70% of the tumor cells. In a recent study, the authors found 56% of oral SCC lacked PTEN protein expression and associated this finding with pathogenesis and carcinogenesis of oral SCC [28].

Our study shows that p-mTOR nuclear expression was significantly related to overall survival. The nuclear import of mTOR has an important role in activating its cytoplasmic signaling. However, the mechanism of nuclear transportation of mTOR and the function of nuclear mTOR remain unclear. Rapamycin is a macrolide antibiotic and an immunosuppressive agent that inhibits mTOR. Its anti-proliferative effect is mediated through the formation of an active complex. Rapamycin also suppresses angiogenesis by decreasing the production of vascular endothelial growth factor [29]. In accordance with our results, a recent study has shown that adverse outcomes are associated with high expression of mTOR in a subset of 72 oral SCCs [30]. Indeed, mTOR inhibitors have shown promising efficacy rates in patients with renal cell and gastric carcinomas [31, 32].

Conclusion

We have evaluated p-AKT, p-mTOR, p-MAPK, PI3K, and PTEN expression by immunohistochemistry and observed that the PI3K/AKT pathway is activated in gingival, hard palate, and alveolar ridge SCCs. Our data show that p-mTOR expression can function as a biomarker of survival in oral SCCs and could be a promising therapeutic target in oral SCC.