Expression of Dysadherin and Cytokeratin as Prognostic Indicators of Disease-free Survival in Patients with Stage I NSCLC

Materials and Methods

Patients, clinical features, and follow-up. The Institutional Review Board approved this study and informed consent for the use of the tumor specimens was obtained either from all the patients or their legal guardians. Tumor samples were obtained from 408 patients with NSCLC who had undergone a surgical resection in this Department between 1994 and 2000. Twenty-four, 59, and 101 of these patients were stage II, III, IV, respectively, and 38 underwent an incomplete resection. The tumor samples from 79 patients were too small to evaluate by immunohistochemical (IHC) staining for the three molecules. A total of 301 patients were excluded from further analyses. Therefore, a total of 107 tumor specimens were evaluated. All of the patients in this series were Japanese, including 71 males and 36 females, with a mean age of 67.8 years (range 47-85 years). There were 34 non-smokers (never smokers), 18 former, and 55 current smokers. Former smokers were defined as those who quit smoking at least 3 years before the time of surgery. The tumor stage was classified according to the Revisions in the International System for Staging Lung Cancer (8). A pathological analysis revealed that 51 patients were at stage IA and 36 at IB. The histological types included 74 adenocarcinoma (69.2%), 22 SQ (21.4%), 6 large cell carcinoma (5.6%), and 3 others. The surgical procedures included a lobectomy in 105 patients, and a pneumonectomy and a segmentectomy in each of the remaining two patients. None of the patients had received either chemotherapy or radiotherapy prior to the resection.

The patients were 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, an analysis of blood chemistry, and measurements of tumor markers. Chest and abdominal computed tomography, brain magnetic resonance imaging, and a bone scintiscan were performed every 6 months for 3 years after surgery. Additional examinations were performed if any symptoms or signs of recurrence were detected. A follow-up was conducted in all patients. The median follow-up period was 55.8 months. Sixty-five patients were alive and free of cancer at the last follow-up, while 19 patients had died of other causes without evidence of cancer, 17 patients had died of cancer, 5 patients were alive with recurrent cancer, and 1 patient had died of other causes with recurrence of cancer. Twenty-two (20.6%) out of the 107 patients demonstrated disease recurrence after surgery.

Detection of E-cadherin and dysadherin expression in tumors. IHC staining was conducted using serial sections from the same paraffin-embedded blocks by previously described methods (4, 9). Briefly, all tissue specimens were formalin-fixed and processed similarly, according to the standard histological methods. A 3 μm-thick formalin-fixed, paraffin-embedded tissue section was treated with 0.3% hydrogen peroxidase for 15 minutes to quench the endogenous peroxidase activity. The sections were then incubated with an E-cadherin mouse monoclonal antibody (C20820; BD Bioscience, San Diego, CA, USA) diluted 1:200, or dysadherin (NCC-M53, kindly provided by Dr Hirohashi, National Cancer Center Research institute Tokyo, Japan) diluted 1:500 (10), respectively, in PBS and incubated at 4°C for at 18 h. The labeled streptavidin-biotin kit (Dako Corporation, Carpinteria, CA, USA) was used as the secondary antibody (11-13). The diaminobenzidine method was used to visualize the peroxidase with hematoxylin as a counterstain. The percentage of immunoreactive tumor cells in five ×400 fields selected randomly on one slide was recorded, and then the final value of positive tumor cells was determined as the average of the number of positively immunostained cells. The stained specimens were scored as previously reported: positive E-cadherin expression (≥70% of tumor cell membranes stained) or negative (<70% of tumor cells stained) (4), and positive dysadherin expression (≥50% of tumor cell membranes stained) or negative (<50% of tumor cells stained) (10). Necrotic areas were excluded from the evaluation. Positive staining of normal epithelial cells adjacent to the tumor and lymphocytes in the primary tumor sections was used as an internal positive control, for E-cadherin and dysadherin staining, respectively. The negative controls were sections stained with the exclusion of the primary antibody. Two observers (K.O. and H.U.), who had no previous knowledge of the clinical parameters and outcomes for each patient, independently reviewed the immunohistochemically stained sections; all discrepancies were resolved by a joint review of the slides in question.

Detection of micrometastasis of cytokeratin-positive cells in the lymph nodes. The lymph nodes were removed from these 107 patients during surgery and they were thereafter analyzed for micrometastasis using cytokeratin IHC staining by previously described methods (4). Briefly, five 4-μm slices, representing every other slice from 10 slices of each paraffin-embedded lymph nodes section, were attached to glass slides. The slides were then stained with primary antibodies against cytokeratin using the labeled streptavidin-biotin method (Dako LSAB kit; Dako Corp). The primary antibodies were mouse biclonal antibody AE1/AE3 (Progen Biotechnik GmbH, Heidelberg, Germany) to cytokeratin. The staining procedures were: (i) deparaffinization with xylene and ethanol, (ii) incubation with the primary antibodies (dilutions: AE1/AE3, 1/200), (iii) incubation with the secondary antibody, (iv) developing with peroxidase-labeled streptavidin and diaminobenzidine-H₂O₂, and (v) counterstaining with hematoxylin. The presence of cytokeratin-positive cells within the whole-body section of the lymph nodes was accepted as evidence of micrometastatic tumor cells, even if only a single cytokeratin-positive cell was detected. Ninety-eight (91.6%) out of 107 patients had undergone dissection of the hilar and mediastinal lymph nodes (systematic nodal dissection) and 9 patients had only the hilar lymph nodes dissected.

Statistical analyses. Statistical significance was evaluated using the Fisher's exact test. A multivariate logistic regression analysis was used to evaluate independent associations. The Kaplan-Meier method was used to estimate the probability of survival, and survival differences were analyzed by the log-rank test. A multivariate analysis was then performed according to Cox's proportional hazards model. The odds ratio and 95% confidence interval were calculated for each variable. Differences with p-values <0.05 were considered to be statistically significant. The data were analyzed using the Stat View software package (Abacus Concepts, Inc., Berkeley, CA, USA).