Prognostic significance of STAT3 and phosphorylated STAT3 in human soft tissue tumors - a clinicopathological analysis

Primary surgical specimens were obtained from 82 patients (51 males and 31 females) who were clinically diagnosed for soft tissue tumors, from Department of General Surgery, Govt. Medical College Hospital, Thiruvananthapuram, India between 2007 and 2008 following approval from the Human Ethics Committee. Of the 82 cases, 48 were malignant, 25 benign, and 9 were of intermediate grade. Tumor stages were classified according to the revised GTNM (grade-tumor-node-metastasis) classification of WHO (2002).

The present study correlated the gross pathological features of soft tissue tumors like tumor size, location, depth, circumscription, encapsulation and presence of necrosis with clinical parameters. Histopathological parameters were studied using 5 μm thick paraffin sections stained with Hematoxylin and Eosin and the tumors were broadly classified into benign, intermediate and malignant.

Resected specimens were fixed with 10% paraformaldehyde and embedded in paraffin blocks. Five-micrometer sections of 82 representative soft tissue tumor blocks were used for immunohistochemical analysis. Sections were deparaffinized in xylene and rehydrated in graded alcohols and water. Endogenous peroxidase activity was blocked via treatment with 2.5% hydrogen peroxide for 20 minutes. Antigen retrieval was performed by placing the slides in boiling citric acid buffer (10 mM sodium citrate and 10 mM citric acid) for 15 minutes. Sections were treated with protein-blocking solution for 30 minutes and primary antibodies such as STAT3 and pSTAT3 (Santa Cruz Biotechnology, Inc, CA) were applied at a 1:100 and 1:50 dilution and incubated overnight at 4°C. After several rinses in phosphate-buffered saline, the sections were incubated in biotinylated secondary antibody for 30 minutes. The bound antibodies were detected by a streptavidin-biotin method, with a Vecta Elite ABC staining kit (Vector Laboratories). The slides were rinsed in phosphate-buffered saline, exposed to diaminobenzidine, and counterstained with Mayer's hematoxylin. For the tumor tissues, nuclear STAT3 and pSTAT3 (Tyr 705) staining were recorded as the numbers of STAT3 and pSTAT3-positive nuclei, divided by the total number of nuclei of at least 10 fields, and then expressed as a percentage. Cytoplasmic positivity of STAT3 and pSTAT3 were measured depending on the intensity of immunoreactivity (independently scored by D.D, AN, and LMR) and scored as mild (+), moderate (++), and intense (+++).

Protein extracts were prepared by homogenizing fresh tissue in lysis buffer comprising 10% NP40, 5 M NaCl, 1 M HEPES, 0.1 M DTT, 0.1 M EGTA, 0.1 M EDTA, protease inhibitors (Sigma) and differential centrifugation (14000 rpm for 10 minutes). The protein concentrations were determined using Bradford's assay and 60 μg of proteins were resolved by 10% SDS-PAGE, and the separated proteins were electrotransferred onto nitrocellulose membrane (Amersham Pharmacia Biotech). After preblocking these membranes with 5% skimmed milk, they were treated with antibodies against STAT3 (1:200, Santa Cruz Biotechnology), pSTAT3 (Tyr 705) (1:200, Santa Cruz Biotechnology), and β- actin (1:5000, Sigma) as primary antibodies and incubated overnight at 4ºC. Horseradish peroxidase-conjugated antirabbit (1:5000, Santa Cruz Biotechnology) and antimouse (1:5000, Santa Cruz Biotechnology) antibodies were used as secondary antibodies and incubated for 1 h at room temperature. Immunoreactive bands were developed with an ECL system (Amersham Pharmacia Biotech, Uppsala, Sweden).

Total RNA was isolated from fresh tissues using TRIzol (Invitrogen) reagent. 10μg of total RNA was converted to cDNA using M-MLV Reverse Transcriptase (Promega) in a 25μl reaction. The relative expression of STAT3 was analyzed using semi-quantitative reverse transcription- PCR with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as an internal control. The primers used were STAT3 (sense), 5'-GGAGGAGTTGCAGCAAAAAG-3'; STAT3 (antisense) 5'-TGTGTTTGTGCCCAGAATGT-3'; GAPDH (sense), 5'-TTGGTATCGTGGAAGGACTCA-3'; GAPDH (antisense), 5'-TGTCATCATATTTGGCAGGTT-3'.The RT-PCR reaction mixture contained 5μl of 10× reaction buffer, 5μl of cDNA template, 0.5 μL each of forward and reverse primers, and 0.5 μL of Dr Taq DNA polymerase (Biogene) in a final volume of 50 μL. The reaction was done at 94°C for 4 min (Initial denaturation), 94°C for 30 s (Denaturation), 60°C for 40 s (Annealing), 72°C for 1 min and 30 s (Extension), and 72°C for 7 min (Final extension) for 35 cycles. Analysis of amplified products was done on 2% agarose gel and visualized using Fluor-S™ MultiImager (Bio-Rad). The PCR products were quantified by densitometric analysis, using Bio-Rad Quantity One software. The mRNA levels of STAT3 were normalized to human GAPDH mRNA levels. A 100-bp ladder was used as a size standard.

Statistical analysis was performed using Intercooled Stata software (Intercooled Stata 8.2 version). The clinicopathological characteristics of the patients were compared between tumor grade, and expression of STAT3 and pSTAT3, using Chi squared or Fisher's exact test. The limit of statistical significance was set at P < 0.05. The effect of clinicopathologic characteristics on STAT3 and pSTAT3 expression were estimated with Odds Ratio (OR) and their 95% Confidence Interval (CI) derived from logistic regression analysis. Sensitivity and specificity of STAT3 and pSTAT3 expression were determined by taking the histopathological grade of tumor as the Gold standard.

The patients included in this study were aged from 1 to 80 years (Mean 42, SD = 19.8). Both age and sex of the patients showed significant association with tumor grade (P = 0.012; P = 0.04). Tumor size and tumor location also showed significant association with grade of the tumor (P = 0.004; P = 0.009). While most of the benign tumors occurred in the extremities (68%), the lower extremities (45.8%) followed by the retroperitoneum (27.1%) were the favored sites for malignant tumors. Tumors of intermediate grade were more common in the trunk (55.6%). Most of the soft tissue tumors in the present study were located in the subcutaneous plane (52.4%) followed by the muscular plane (28%).

Among the 82 tumors studied, 38 were well-circumscribed and showed significant association with tumor grade (P < 0.001). Necrosis was studied in all the tumors and significant association was observed with the grade of the tumor (P < 0.001). Tables 1 list the clinicopathological characteristics of the soft tissue tumors selected for the study. Pathologic features of the representative benign, intermediate and malignant soft tissue tumors were given in Figure 1.

Expression of STAT3 and pSTAT3 showed statistically significant association with histopathological parameters as evidenced by Chi squared and Fisher's exact test [See Additional file 1 Table S1]. STAT3 and pSTAT3 expressions were significantly associated with grade of the tumor (P < 0.001). Malignant tumors were 107.3 times more likely to express STAT3 (OR = 107.3, 95% CI: 20.24-569), and 7.5 times more likely to express pSTAT3 (OR = 7.5, 95% CI: 2.28-24.5) when benign or intermediate tumor is the reference [Table 3]. The sensitivity and the specificity of STAT3 were 95.8% and 76.5% and pSTAT3 were 50% and 88.2%, respectively, with histopathological grade. In addition, Table 4 represents the association between clinicopathologic characteristics and expression of STAT3 in malignant soft tissue tumors.