We enrolled 78 patients with primary OSCC who were treated in the Department of Oral and Maxillofacial Surgery at Kyushu University Hospital from 2005 to 2011. The average age of the patients was 65.7±13.2 years (range, 19–89). Fifty-two patients were males and 26 were females. An informed consent was obtained from all the participant patients before any procedure or treatment was initiated (IRB no.: 25-227). Thirty-nine patients with locoregionally advanced OSCC underwent chemoradiotherapy preoperatively. Following the initial biopsy, all the specimens were fixed in 4% buffered formalin solution and were embedded in paraffin blocks. Subsequently, the paraffin-embedded specimens were processed to 5 μm thick sections, stained with hematoxylin and eosin (H&E) and examined by experienced oral pathologists to confirm the diagnosis and histologic grade. The tumor stage was classified according to the TNM classification of the International Union Against Cancer (21). Tumor histologic grade was defined according to the WHO classification (22). The mode of the tumor invasion was determined from H&E stained specimens according to the Yamamoto-Kohama criteria as follows: grade 1 = well-defined borderline; grade 2 = cords, less-marked borderline; grade 3 = groups of cells, no distinct borderline; grade 4 = diffuse invasion (4C = cord-like type; 4D = widespread type) (23). Patients and tumor characteristics are shown in Table I.

The patients with locoregionally advanced OSCC received external beam irradiation to the primary tumor and the metastatic lymph nodes in daily fractions of 2 Gy, 5 times weekly, for 3 weeks. S-1 (TS-1; Taiho Pharmaceutical, Tokyo, Japan), an oral fluoropyrimidine preparation that consists of tegafur, 5-chloro-2,4-dihydroxypyridine (a dihydropyrimidine dehydrogenase inhibitor) and potassium oxonate which inhibits orotate phosphoribosyltransferase in the gastrointestinal tract, was used as an anticancer drug in this regimen. Oral administration of S-1 started one week prior to the radiotherapy and was continued throughout the radiotherapy period. Standard individual doses of S-1 were calculated according to the body surface area (BSA): BSA <1.25 m², 80 mg; 1.25 ≤BSA <1.5 m², 100 mg; BSA ≥1.5 m², 120 mg. However, in patients with reduced renal function (decreased creatinine clearance values), S-1 was administered at a lower dose, generally one step lower than the standard dose. Radical surgery was carried out at 2–6 weeks (average, 27.5±5.22 days) after the end of the preoperative chemoradiotherapy.

Immunohistochemical staining was performed on 5 μm thick sections that were sliced serially from paraffin-embedded blocks after formalin fixation of the excised specimens. The sections were deparaffinized in xylene and rehydrated in graded ethanol (100, 95, 90, 85 and 75%). For antigen retrieval, the sections were immersed in Dako Target Retrieval Solution (Dako Cytomation, Denmark) and autoclaved at 121°C for 5 min. The endogenous peroxide activities were then eliminated with 3% hydrogen peroxide for 5 min, and the sections were rinsed twice for 10 min with phosphate-buffered saline (PBS) at pH 7.4. Non-specific protein bindings were attenuated by incubation for 30 min with 10% goat serum and then the sections were incubated with each primary antibody overnight at 4°C. The following antibodies were used: anti-human monoclonal IL-6 antibody (Abcam, UK; diluted 1:50), anti-human monoclonal IL-6 receptor (IL-6R) antibody (Invitrogen, CA, USA; diluted 1:50) and anti-human monoclonal phospho-signal tranducer and activator of transcription 3 (p-STAT3) antibody (Cell Signaling Technology, MA, USA; diluted 1:400). The sections were rinsed twice for 10 min with PBS and incubated with secondary antibody conjugated with peroxidase-labeled amino acid polymer for 1 h at room temperature. After being rinsed with PBS twice for 10 min, the immunoreactivity was visualized by immersing the sections in 3,3′-diaminobenzidine and 0.6% hydrogen peroxide (DAB substrate kit; Nichirei, Tokyo, Japan). Subsequently, the sections were counterstained with Mayer’s hematoxylin, dehydrated in graded ethanol (75, 85, 90, 95 and 100%), cleared with xylene and finally mounted with permanent mounting medium (Malinol mounting medium; Muto Pure Chemicals, Tokyo, Japan). Negative controls were prepared by substituting PBS for each primary antibody. To evaluate the expression of IL-6 in the OSCC, positively stained cancer cells were counted in at least three randomly selected areas at ×200 magnifications, and then each percentage of these positive cancer cells was calculated as labeling index (LI). The LI was computed by dividing the number of the positively stained cells by that of all the cancer cells. The patients with OSCC were divided into three groups as follows: negative group = LI <5%; low IL-6 group = 5% ≤ LI <30%; high IL-6 group = LI ≥30%.

The classification of therapeutic efficacy established by Shimosato et al (24) was used to evaluate the histopathological response of the primary site tumors: grade 0, no noticeable change; grade I, minimal cellular changes, but the majority of the tumor cells appear viable; grade IIa, despite the presence of the cellular changes and partial destruction of the tumors, the tumor is still readily recognizable and many tumor cells appear viable; grade IIb, tumor destruction is extensive, but viable cell nests are present in small areas of the tumor (up to one quarter of the tumor mass, excluding areas of coagulative necrosis); grade III, only a few scattered, markedly altered and presumably non-viable tumor cells are present, singly or in small clusters, and few or no viable cells are seen; and grade IV, no tumor cells remaining in any section. Cutting of resected specimens was carried out by step-section method at intervals of 5 mm.

All statistical analyses in the present study were performed by JMP software version 11 (SAS Institute, NC, USA). Chi-squared test was used to assess the significant differences between each group. Survival rates were also calculated by Kaplan-Meier method and the P-value was calculated by Log-rank test. A P-value of <0.05 was considered to be statistically significant.

Immunoreactivity for IL-6 was almost not detected in the adjacent non-malignant oral epithelia and weakly investigated in the submucosal tissues. In the OSCC patients, the IL-6 expression pattern was different in the individual specimens. The number of patients in the negative, low and high expression group were 24, 22 and 32, respectively. The OSCC patients were thus divided into three groups according to IL-6 positive rates as follows: negative, low and high expression groups. In the negative group, most cancer cells were negative for a IL-6 immunoreactivity whereas stromal cells around the cancer nests expressed IL-6 slightly. IL-6 expression of the low expression group was scattered in the cancer cells whereas almost all the cancer cells expressed IL-6 in the high expression group. The stromal cells also strongly expressed IL-6 in the high expression group (Fig. 1).

In order to examine whether IL-6 signaling is activated in the OSCC cells, the expression of IL-6R and p-STAT3 was further examined immunohistochemically in the patients of the high IL-6 expression group. IL-6R was detected in the cytoplasm and plasma membrane of almost all the cancer cells and observed in harmony with that of IL-6 in almost all the OSCC cells of the IL-6 high expression group (Fig. 2A and B), but not or weakly in the negative and low expression group (data not shown). p-STAT3 expression was localized in the nucleus of the cancer cells and scattered widely in the cancer nests (Fig. 2C).

The associations of IL-6 expression with the clinicopathologic factors of the OSCC patients were examined. The patients in the high expression group had a more advanced clinical stage than those in the negative and low expression groups. Furthermore, the prevalence of the cervical lymph node or distant metastasis in the high expression group was significantly higher than those in the negative and low expression groups (P<0.05, Chi-squared test). On the contrary, other clinical factors including gender, primary site, clinical T classification, local recurrence rate, histologic grade and mode of invasion did not show significant differences among these groups (Table I).

In the 39 patients who underwent preoperative chemoradiotherapy for locally advanced OSCC, the association of IL-6 expression in the cancer cells with histologic tumor response in the resected specimens was further examined. Equal to or higher than Grade IIb were judged to be histologically effective. All the patients in the negative group were good responders, whereas only a third of the patients had a histologically effective response in the high expression group. Increased IL-6 expression in the cancer cells was significantly associated with poor response to preoperative chemoradiotherapy (^*P<0.05, ^**P<0.01, Chi-squared test Table II).

In order to examine whether STAT3 signaling is also activated in the residual cancer cells after chemoradiotherapy, the expression of IL-6R and p-STAT3 was further examined immunohistochemically in the 12 patients of the high expression group with poor response to chemoradiotherapy. In all the cases, IL-6R was detected in the cytoplasm and cytomembrane of almost all residual cancer cells, while p-STAT3 was expressed only in the outer layers of the cancer nest (Fig. 3).

For the purpose of evaluating the correlation between immunoreactivity for IL-6 in cancer cells and the clinical outcomes of the patients with OSCC, the survival rates were calculated by the Kaplan-Meier method. In the cause-specific cumulative survival curves, the patients in the high expression group had a significantly more unfavorable outcome than those in the negative and low expression groups (Log-rank test, P<0.05; Fig. 4). The cumulative survival rate for 5 years in the high expression group was 68.0%, whereas that in the negative and low expression groups was 90.0%.