Carbonic anhydrase IX in tumor tissue and sera of patients with primary cervical cancer

Tumor tissue from 221 patients undergoing surgery for primary cervical cancer at 4 different institutions between 1993 and 2008 was analyzed. Detailed patient characteristics are listed in Table 1. In addition, 46 preoperative serum samples from a subset of the included patients were analyzed. Serum samples were not routinely taken from all included patients before surgery but on a voluntary basis in one of the participating centers. To assure the analysis of a representative collective, FIGO stages were compared to those of the whole cohort showing a similar stage distribution (FIGO stage Ib1 52.2%, Ib2 13.0%, IIa 6.5%, IIb 17.4%, III 0.0%, IVa 4.3%, IVb 6.5%, compare Table 1). Clinicopathologic factors were evaluated by reviewing medical charts and pathologic reports. For tumor staging, 6^th edition UICC TNM classification and stage groupings were used [30]. Informed consent for the scientific use of patient data, tumor tissue and serum had been obtained from all patients in coordination with the local ethics committee according to the principles of the declaration of Helsinki (Ethics committee of the Medical Board Hamburg, reference number #190504). In the current study, all data were analyzed anonymously.

For 180 patients detailed therapy information and for 175 patients detailed follow-up data from the date of primary surgery to the date of death or last contact (July 2009) were available. The treatment of cervical cancer during the investigational period consisted of radical hysterectomy and resection of the pelvic and paraaortic lymph nodes via laparotomy for FIGO stages I and II. Only in stage Ia and Ib disease with fertility preserving objective radical hysterectomy was omitted. Adjuvant radio/chemotherapy of the pelvis and the paraaortic region was performed in case of advanced disease according to effective German Guidelines [31]. For FIGO stage III and IV cervical cancer primary radio/chemotherapy was generally recommended. In our study, 24 patients were included with FIGO stage III/IV disease who received primary surgery. Those patients underwent surgery either to prevent further tumor invasion of the bladder and/or rectum or were not classified as FIGO III/IV until surgery (predominantly in case of paraaortic lymph node metastases).

To analyze the tissue-expression of CAIX, we performed immunohistochemistry (IHC) with paraffin embedded tumors from all included patients, using a commercially available CAIX IHC kit (Oncogene Science, Cambridge, Mass 0214, USA) according to the manufacturer's instructions. Tissue sections were cut at a thickness of 5 μm and mounted on slides (Superfrost/Plus, Sondheim Germany), dewaxed with Xylene and gradually hydrated. Afterwards a peroxidase blocking solution was applied and sections were incubated at room temperature for 5 minutes followed by several washings in phosphate buffered saline (PBS) (CA IX IHC Kit, Reagent H). Adjacent all sections were probed with the primary mouse anti-CA IX antibody (prediluted at 0.5 μg/mL, CA IX IHC Kit, Reagent B) incubating them at room temperature for 30 minutes. After several washings in PBS, slides were covered with the biotinylated secondary antibody (CA IX IHC Kit, Reagent C) for 10 minutes, washed several times and incubated with HRP-labeled streptavidin (CA IX IHC Kit, Reagent D) for another 10 minutes to be washed three times again. Finally the DAB substrate-chromogen solution (CA IX IHC Kit, Reagent F1, F2 and F3) was applied for 5 minutes followed by washings in Aqua dest. The slides were briefly counterstained with haematoxylin (CA IX IHC Kit, Reagent G) and dehydrated before mounting.

The IHC staining was evaluated independently by two gynecopathologists using the immunoreactive score (IRS) according to W. Remmele and H. E. Stegner as product of staining intensity (graded 0-3) and percentage of positive cells (graded 0-4, 0: 0%, 1: <10%, 2: 10-50%, 3: 51-80%, 4: >80%) resulting in a score from 0-12 [32].

CAIX was quantified by a commercially available ELISA (Oncogene Science, Cambridge, Mass 0214, USA). Control samples consistent of purified CAIX protein in buffer with low, medium and high concentrations as included in the ELISA kit ("MN/CAIX ELISA controls"). Serum samples and controls were diluted 1:2 with sample diluting buffer (containing bovine serum albumin, mouse IgG, buffer salts and 0.09% sodiumazide). One hundred microliters of the standards, of diluted control samples and of diluted serum samples were dispensed into the wells of a 96-well plate (coated with the monoclonal capture antibody) and incubated for two hours at room temperature on a shaker at 800 rpm. Wells were washed, and 100 μl of the detection antibody (containing biotinylated anti-CAIX antibody) was added. The plates were incubated for 30 min at room temperature, washed, and then further incubated with 100 μl of a streptavidin-HRP (horseradish peroxidase) conjugate for 30 min at room temperature. After washing, 100 μl of chromogenic substrate (TMB blue substrate) was added for 30 min at room temperature. The reaction was stopped with 100 μl of 2.5 N sulphuric acid and absorbance was read at 450 nm by an automated plate reader (Tecan, Crailsheim, Germany). The CAIX concentration was estimated from the standard curve. Each sample, standard, and control was analyzed in duplicate.

All statistical analyses were conducted using the R open source statistical software package [33]. p-values < 0.05 were considered statistically significant. In accordance with previously published data on CAIX expression in epithelial tumors we analyzed CAIX expression in a two class system, merging IRS score 0 and 1 into one class corresponding with none/weak CAIX expression levels and IRS score 2 - 12 into another class corresponding with moderate/strong intratumoral CAIX expression [19, 34, 35]. Mann-Whitney-U-Tests and Fisher's exact tests were used to calculate correlations between CAIX expression, serum CAIX and clinicopathological parameters. Kaplan-Meier analyses of overall and disease-free survival were performed. The log rank test was applied to examine the relationship between serum CAIX, intratumoral CAIX expression and survival. Additionally, we performed a stratified log rank test, comparing CAIX effects on survival in patients with FIGO stage I and FIGO stage II-IV disease, to investigate potential stage dependant differences. The Cox Proportional Hazard Model was used for multivariate analyses.

A total of 221 patients were included in this study; detailed characteristics are listed in Table 1. For 180 patients detailed therapy information and for 175 patients detailed follow-up data were available. All patients underwent radical surgery. No further treatment was initiated in 88 patients (48.9%), whereas 92 (51.1%) patients received adjuvant radio- and/or chemotherapy. The predominantly applied chemotherapy regimen was cisplatin as a single agent, 4 patients received a cisplatin based combination therapy. Median follow-up time was 49 months. At the time of last follow-up, 131/175 (74.9%) patients were alive and without relapse whereas 38/175 (21.7%) patients had died from the disease. In case of disease recurrence (n = 44), relapse was locoregional in 10 patients, distant in 13 patients and both in 18 patients (3 cases unknown).

A total of 221 tumor samples were stained for determination of CAIX expression. CAIX was detectable in 181 (81.9%) of all specimens. 137 (62.0%) samples exhibited a moderate or strong (score 2-12) staining result (Table 1). Figure 1 shows representative tumor samples with none/weak/moderate/strong CAIX expression. CAIX protein expression was always membranous. Tumor associated stromal cells as well as inflammatory cells stained negative for CAIX. Table 2 summarizes the correlation of CAIX expression with clinicopathological parameters. Moderate/strong expression was associated with squamous histology (p = 0.024), advanced FIGO and pT stage (p = 0.013 and p = 0.001), greater invasion depth (p = 0.025), undifferentiated tumor grade (p < 0.001) and high preoperative SCC-Ag concentration (p = 0.042). Furthermore, patients with moderate/strong intratumoral CAIX expression had a higher number of metastatic lymph nodes compared to those with none/weak intratumoral expression levels (p = 0.047). Fifteen (11.0%) of the patients with moderate/strong intratumoral CAIX expression showed three or more lymph node metastases whereas in only 2 cases (2.4%) with none/weak CAIX expression three or more positive lymph nodes were diagnosed. We could not demonstrate an association between CAIX expression and lymphovascular space invasion or patients' age. There was an association between none/weak intratumoral CAIX expression and longer survival; however, this result failed to reach statistical significance (overall survival p = 0.118, Figure 2; disease-free survival p = 0.220, not shown). The potential negative effect of a moderate/strong intratumoral CAIX expression on survival did not significantly differ between FIGO stages I and II-IV in a stratified survival analysis (disease-free survival p = 0.580, overall survival p = 0.331, not shown).

Median preoperative serum CAIX was 104 pg/mL (mean 137.0, range 23-499 pg/mL). Patients with a moderate/strong intratumoral CAIX expression did not show significantly higher serum concentrations (median 110 pg/mL, range 23-499 pg/mL) compared to patients with none/weak expression (median 91 pg/mL, range 27-257 pg/mL, p = 0.466). Furthermore, serum CAIX did not show any association with the analyzed clinicopathological factors (tumor stage, lymph node involvement, lymphovascular space invasion, invasion depth, tumor grade, histological type, preoperative serum SCC-Ag). Nevertheless, there was a non-significant association between high preoperative serum CAIX and shorter disease-free survival (disease-free survival p = 0.127, Figure 3; overall survival p = 0.320, not shown). After stratifying for FIGO stages, the negative effect of a serum CAIX above median on survival was especially distinct in patients with FIGO stage I disease (disease-free survival p = 0.051, Figure 4; overall survival p = 0.059, not shown)

The independent prognostic impact of intratumoral CAIX expression and preoperative serum CAIX for disease-free and overall survival adjusted for tumor stage (FIGO), pN, tumor grade and depth of invasion was analyzed using the Cox Proportional Hazard Model. Higher FIGO stage and nodal involvement were shown to be independent prognostic factors for decreased survival. Neither moderate/strong CAIX expression nor serum CAIX above median before surgery could be demonstrated independently relevant for prognosis in this patient cohort.