TMPRSS4 Expression as a Marker of Recurrence in Patients with Lung Cancer

Materials and Methods

The Institutional Review Board of our University approved this study (H25-185).

Case selection for the microarray. Tumor samples were obtained from 668 patients with primary lung cancer who had undergone surgical resection between 2005 and 2011 at our Department. Two hundred and thirty-five of these patients had double cancer, including multiple lung cancer, 38 of whom underwent incomplete resection. Thirteen and seven cases were positive and suspicious on washing cytology, respectively. Furthermore, the number of cases with a carcinoembryonic antigen (CEA) level of more than 2.5 ng/ml and those treated with limited resection was 112 and 21, respectively. As a result, 426 patients were excluded from further analyses. Therefore, 242 tumor specimens were qualified as candidates for the evaluation. Among them, we selected two available cases with frozen specimens followed-up for a three-year period after surgery by the same operator over the same period to avoid selection bias. The tumors were derived from one case without recurrence for a long-term survivor with advanced disease after resection (Figure 1) and a case of recurrence despite pathological stage IA disease (Figure 2).

Cell line selection for the microarray. A human lung adenocarcinoma cell line, A925L, established from a surgical specimen obtained from a male Japanese patient (T2N2M0, stage IIIA), was maintained in RPMI-1640 medium, supplemented with 10% fetal bovine serum (FBS), penicillin (100 U/ml) and streptomycin (50 g/ml), in a humidified CO₂ incubator at 37°C (3). A925LPE3 cells have a high potential to produce thoracic tumors and pleural effusion in mice. A925L and A925LPE3 harbor an EML4–ALK gene fusion (variant 5a) and are sensitive to ALK-TKIs. We used A925L and A925LPE3 cell lines as a model for cells with parental and highly metastatic potential, respectively. We previously established in vivo imaging models for pleural carcinomatosis, bone metastasis and brain metastasis (3).

Microarray expression analysis. Total RNA was extracted from each serum sample from the cases and cell lines as stated before using “3D-Gene” RNA extraction reagent obtained from a liquid sample kit (Toray Industries, Inc., Tokyo, Japan) using a combination of formalin-fixed, paraffin-embedded (FFPE) tissues and cell lines. Total RNA was also obtained from sections in accordance with the manufacturer's instructions. Hybridization was performed using the supplier's protocols (www.3d-gene.com). The amplified RNA was hybridized for 16 hours to the highly sensitive DNA microarray, 3D-Gene Human Oligo chip 25 k, ver. 2.1 (Toray Industries, Tokyo, Japan), which permits the detection of 24,460 mRNAs. The DNA microarray product was washed according to the manufacturer's instructions, followed by image scanning using a 3D-Gene® Scanner 3000 (Toray Industries) and data processing using 3D-Gene® Extraction 2.0.0.4 (Toray Industries). At a laser power setting of 60%, the photomultiplier sensitivity was adjusted so that the fluorescence signal intensity from the two fluorophores, Cy3 and Cy5, at the gain control spots on the microarray was comparable. The ratio of Cy3 and Cy5 signal intensity (Cy3/Cy5, R) was log translated and used as a calibration factor [Log2(R)] to standardize the signal intensity (4). The mean intensity and standard deviation (SD) of the background signal were calculated using the signal intensity of the blank spots residing within the 95% confidence interval (95% CI). Genes with a background signal greater than 2×SD were considered to exhibit positive expression. The mean background level was subtracted from the signals of the detected genes (5, 6). In this study, the mRNA expression profiles of the following one paired-clinical condition were statistically compared: i) the case without recurrence versus the recurrent case, and ii) the cell line with parental versus that with highly metastatic potential.

Patients, clinical features and follow-up. Tumor samples were obtained from 410 patients with primary lung cancer who had undergone surgical resection between 2003 and 2007 at our Department. The number of patients with lung adenocarcinoma was 281. Five of these patients had stage IV disease and 28 underwent incomplete resection. The tumor samples from 87 patients were too small to be evaluated with immunohistochemical (IHC) staining to determine the TMPRSS4 status. As a result, these 120 patients were excluded from the further analysis. Therefore, 161 tumor specimens were evaluated. All of the patients were Japanese, consisting of 90 males and 71 females in this series, with a median age of 71 years (range=23-88 years). There were 67 never smokers, 45 former smokers and 49 current smokers. The former smokers were defined as those who had quit smoking at least three years before the time of surgery. The tumor stage was classified according to the TNM Classification for Lung Cancer (7). Based on the pathological stage, 98 patients had tumors of stage IA, 30 patients had tumors of stage IB, seven patients had tumors of stage IIA, 10 patients had tumors of stage IIB, 13 patients had tumors of stage IIIA and three patients had tumors of stage IIIB. Twenty-one (13.0%) patients received adjuvant chemotherapy, as follows: carboplatin plus paclitaxel (n=13), carboplatin plus gemcitabine (n=6) and tegafur–uracil (n=2) (8).

The patients were generally followed-up every month within the first postoperative year and at approximately 2- to 4-month intervals thereafter. The evaluations included a physical examination, chest roentgenography, analysis of blood chemical parameters and measurements of the tumor marker levels. Chest and abdominal computed tomography, brain magnetic resonance imaging and bone scintiscan assessments were performed every 6 months for 3 years after surgery. Additional examinations were performed if any signs or symptoms of recurrence were detected. Follow-up was conducted for all patients. The median follow-up period was 61.0 months.

IHC staining of FFPE tumor samples. IHC staining was carried out using serial sections obtained from the same FFPE blocks according to previously described methods (9). Briefly, all tissue specimens were fixed in formalin and processed similarly, according to standard histological practices. A 3-μm-thick FFPE tissue section was prepared from each specimen. All specimens were stained with hematoxylin–eosin for a histological diagnosis. The sections were briefly immersed in citrate buffer [0.01 mol/l citric acid (pH 6.0)] and then incubated twice for 10 min at 121°C in a high-pressure sterilization oven for antigen retrieval. The sections were then incubated with TMPRSS4 antibody (Proteintech: 11283-1-AP; Proteintech Group, Inc., Chicago, IL, USA) diluted at 1:50 in phosphate-buffered saline overnight at 4°C (10).Thereafter, IHC staining was performed using the labeled polymer method (Histofine Simple Stain MAX-PO kit; Nichirei, Tokyo, Japan) according to the manufacturer's instructions (9). The positive control for TMPRSS4 was the colon cancer specimens. The negative control used rabbit IgG (Dako, Glostrup, Denmark) instead of the primary antibody.

Evaluation of the stained specimens. Following IHC detection of the protein expression in each specimen, the percentage of immunoreactive tumor cells in five ×400 fields selected randomly on one slide was recorded, and the final count of positively stained tumor cells was determined as the average number of positively immunostained cells. Initially, four groups were assigned for the assessment of proportional scores for positive staining according to the frequency of positively stained tumor cells (0, none; 1, <25%; 2, <25-50%; 3, >50%). In order to determine the presence of any correlations with the clinicopathological characteristics, the protein expression scores were divided into positive and negative groups. The TMPRSS4 expression status in the cytoplasm of the tumors was categorized as negative when the score was 0-2 and positive when the score was 3. The slides were examined independently by two of the investigators (Y.C. and Y.K.), who were blinded to the patients' clinicopathological data. The concordance rate was 87.1%. When discrepancies were found between the two investigators, a consensus was reached via simultaneous examination by both investigators using a double-headed microscope.

Statistical analysis. Statistical significance was evaluated using the Chi-square test or Fisher's exact test. The Kaplan–Meier method was used to estimate the probability of survival, and survival differences were analyzed according to the log-rank test. Univariate logistic regression models were used to evaluate independent associations.

The odds ratio (OR) and 95% CI were calculated for each variable. Differences were considered to be statistically significant for p<0.05. The data were analyzed using the Stat View software package (Abacus Concepts, Inc., Berkeley, CA, USA).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

A: Case 1: Case without recurrence: Chest computed tomography (CT) showing a tumor in the upper left lobe. B and C: Chest CT shows mediastinal lymph node swelling. Left upper lobectomy and and systematic lymphadenectomy were performed. The pathological findings showed papillary carcinoma (moderately differentiated adenocarcinoma), with ly3, v1, n2 (#4:1/5, #5:1/3, #6:4/5, #7:0/2, #10:0/3, #11:0/1, #12u:0/2, #12l:0/1). The pathological stage was judged to be T1N2M0 stage IIIA. EGFR was judged to be wild-type. Adjuvant therapy was administered. The patient has remained alive for five years without evidence of recurrence.