Validation study of direct sequencing and PCR-invader for detecting EGFR mutations in non-small cell lung cancer (NSCLC)

8094

Background: EGFR mutation is both a predictive and prognostic factor for NSCLC treated with EGFR-TKI. Although new, highly sensitive methods for detecting EGFR mutations are currently available, these methods have not been validated. Methods: To validate direct sequencing and PCR-invader for detecting EGFR mutation, we analyzed 124 NSCLC by both methods concomitantly. Tumor tissues were obtained by surgical resection. Formalin-fixed paraffin-embedded specimens were prepared to analyze EGFR mutation. In direct sequencing, Exon 18, 19, and 21 of the EGFR gene were amplified, and PCR amplification products were sequenced directly (Mitsubishi Chemical Medience Corporation). PCR-invader was performed using the invasive cleavage of probe oligonucleotides to detect 10 mutations including Exon 18, 19, 20, 21 (BML incorporation). Results: EGFR mutations were detected in 51 patients (41%) by direct sequencing and 56 (45%) by PCR-invader. Discordance between two methods was found in 12 patients (10%). Exon 19 deletion was detected in 18 and 22 patients respectively. Exon 21 L858R was detected in 30 and 32 patients respectively. Each mutation in exon 19 deletions or L858R detected by direct sequencing could also be identified by PCR-invader. Overall 45 mutations were concordant by both methods. In two patients who received gefitinib, one patient with wild type by both methods did not respond to gefitinib. On the other hand, the other patient expressing Exon 19 deletion by PCR-invader but regarded as wild type by direct sequencing responded to gefitinib monotherapy. Conclusions: Discrepancy between two methods for detecting EGFR mutation was demonstrated and PCR-invader seems to be more sensitive. Further investigation including other highly sensitive methods is currently underway.

No significant financial relationships to disclose.

Interstitial lung disease associated with amrubicin in the treatment of patients with small cell lung cancer

e19018

Background: Amrubicin is a novel anthracycline agent with significant activity against small-cell lung cancer (SCLC). The adverse pulmonary effects of amrubicin are less well known. We investigated the incidence of acute interstitial lung disease (ILD) in SCLC patients receiving amrubicin. Methods: Medical records were used to retrospectively investigate a total of 100 SCLC patients treated with single-agent amrubicin at the National Cancer Center Hospital East between June 2003 and March 2008. The radiographic records and clinical data of these patients were reviewed to find patients who developed acute ILD after being treated with amrubicin. Results: Seven of 100 SCLC patients subsequently developed pulmonary infiltrates without underlying heart disease after receiving amrubicin, and were identified as cases of acute ILD associated with amrubicin. The median time from the start of amrubicin treatment to the development of ILD was 22 days (15–94 days). Among 7 patients who developed ILD, 6 patients received treatment with corticosteroids and ILD improved in 3 patients. However 3 patients died of respiratory failure. The incidence of ILD was 33% (4/12) in patients with preexisting pulmonary fibrosis and 3% (3/88) in patients without preexisting pulmonary fibrosis, and the difference between the two groups was statistically significant (P = 0.0036). Conclusions: Our findings indicate that amrubicin can cause severe ILD, and that the incidence of ILD might be higher in the patients with preexisting pulmonary fibrosis. Physicians need to be alert to the possibility.

No significant financial relationships to disclose.

Analysis of optimal timing of gefitinib in sensitive non-small cell lung cancer (NSCLC) patients: Should we use gefitinib as first-line chemotherapy?

8068

Background: In gefitinib-sensitive NSCLC such as Asian origin, adenocarcinoma, and never/light smoker, the superiority of gefitinib relative to carboplatin/paclitaxel as first-line chemotherapy was recently demonstrated (IPASS). However, the optimal timing of this agent is still controversial. Methods: To assess the optimal timing of gefitinib for sensitive NSCLC, we retrospectively reviewed 720 NSCLC patients treated with this agent at National Cancer Center Hospital East between 2002 and 2008. Results: A total of 190 patients responded to gefitinib monotherapy (response rate 26%) were analyzed in this study. Forty-four patients (23%) were treated with gefitinib as first-line chemotherapy, and 146 (77%) as second or more-line chemotherapy. Patient characteristics (first-line/second or more-line) were as follows: median age (range) 68 (44–82)/64 (33–82); female 84/67%; non-smoker 80/57%; PS0–1 84/74%; adenocarcinoma 93/97%; stage IV 43/54%; recurrence after surgical resection 45/26%. There were some imbalances between the two groups in terms of female (p=0.0296) and recurrence after surgical resection (p=0.0187). Although 64% of second or more-line group received platinum-based chemotherapy in their clinical course, only 27% of first-line group received it. Median follow-up time was 17.2 months. Progression free survival (PFS) from the start of gefitinib was similar between first-line and second or more-line groups (median PFS: 11 vs. 11 months, p=0.3085). However, overall survival (OS) from the start of first-line chemotherapy was significantly longer in second or more-line group (median OS: 23 vs. 33 months, p=0.0298). Conclusions: Even if gefitinib is administered to sensitive patients as first-line or second or more-line chemotherapy, it should have similar efficacy. The significantly longer survival in second or more-line group may be on account of more percentage of patients receiving platinum-based chemotherapy.

No significant financial relationships to disclose.

Chemoradiation therapy in patients (pts) with small cell lung cancer (SCLC) with pericardial effusion but no distant metastasis

7555

Background: Our previous retrospective analysis demonstrated that the survival of the limited-disease (LD) SCLC pts with ipsilateral pleural effusion was intermediate between those of LD pts without ipsilateral pleural effusion and extensive-disease (ED) pts, and that long-term survival was achieved by LD-SCLC pts with ipsilateral pleural effusion who successfully underwent chemoradiotherapy (CRT) (J Thorac Oncol 2008;3:723–7). We retrospectively investigated the clinical course and outcome in pts with SCLC with pericardial effusion but no distant metastasis and examined the overall survival in pts who received chemotherapy and definitive thoracic radiotherapy (TRT). Methods: The medical records of SCLC pts who received treatment at the National Cancer Center Hospital East between July 1992 and December 2007 were reviewed. During this period 767 pts were newly diagnosed as having SCLC. Four-hundred seventeen pts had no distant metastasis. Ninety-six of those 417 pts (23%, 95% confidence interval (CI): 19–27%) had pleural or pericardial effusion or disseminated pleural nodules, and were included in this study. The 96 pts were divided into two groups: group A included pts with pericardial effusion (n=33), and group B included pts who had pleural effusion and/or disseminated pleural nodules, but did not have pericardial effusion (n=63). Sixteen pts had both pleural and pericardial effusion. Results: All but one patient received systemic chemotherapy. A remaining patient with pleural effusion received only best supportive care. In group A, 19 pts received chemoradiotherapy. TRT was conducted concurrently with 3 or 4 cycles of chemotherapy in 12 pts and sequentially in 7 pts. The response rate for first-line chemotherapy was 79%. In group B, 26 pts received chemoradiotherapy. Survival data were shown as below. Conclusions: Long-term survival was seldom achieved by SCLC pts with pericardial effusion but no distant metastasis, even if they underwent chemoradiotherapy.

[Table: see text]

No significant financial relationships to disclose.

Trends in chemotherapy for elderly patients with non-small cell lung cancer: The experience of one institution

e19081

Background: In approximately the year 2000, the results of a number of important studies of non-small-cell lung cancer (NSCLC) were published. These included the first elderly-specific phase III trial and phase III trials showing a survival benefit for docetaxel when used as second-line chemotherapy. These results had a significant impact on clinical practice relating to advanced NSCLC in elderly patients. Methods: Between July 1992 and December 2003, 223 patients with NSCLC aged ≥70 years received chemotherapy alone as their initial treatment at the National Cancer Center Hospital East. These patients were divided into 2 groups: those that began treatment between 1992 and 1999 (group A) and between 2000 and 2003 (group B). The details of chemotherapy regimens and outcomes were compared. Results: In group A, 83% of patients received platinum-based chemotherapy, two-thirds of these regimens comprised platinum[[Unsupported Character -  ]]plus second-generation combination chemotherapy. In contrast, although 55% of patients received platinum-based chemotherapy in group B, 41% of patients received non-platinum-based chemotherapy. Among patients in group B, performance status was significantly associated with the selection of platinum-based or non- platinum-based chemotherapy; age was marginally associated with this selection. In group A, second-line chemotherapy was administered to only 4 patients (5%) and no patients received third-line chemotherapy. In group B, second-line and third-line chemotherapy was administered to 62 (42%) and 22 (15%) patients, respectively. Median survival time (MST), 1-year survival rate, and 2 year-survival rate were 6.7 months, 14%, and 7%, respectively, in group A, and 8.1 months, 35%, and 20% in group B (p = 0.0109). However, these differences disappeared when patients treated with epidermal growth factor receptor tyrosine kinase inhibitor (EGFR-TKI) were removed from the analysis (6.7 months in group A vs. 6.9 months in group B, p = 0.3684). Conclusions: In and after the year 2000, chemotherapy regimens changed greatly and survival of elderly patients significantly improved. However, the improvement in survival seems to be due mostly to the use of EGFR-TKI treatment.

No significant financial relationships to disclose.

Stage-size relationship in long-term repeated CT screening for lung cancer: Anti-Lung Cancer Association project

1540

Background: We have investigated the individualized benefit of CT screening as Anti-Lung Cancer Association projects (presented at ASCO 2006–2008). However, there has not been enough information about the relationship of lung cancer stage to tumor size in repeated CT screening. Therefore, we evaluated the stage-size relationship of these asymptomatic lung cancer cases diagnosed by long-term repeated screening with low-dose helical CT. Methods: Among 2,120 participants (mean age 63 years, 87% male, and 83% smoker) at baseline screening, 1,877 underwent semiannually repeated screening from 1993 to 2004. Nineteen (0.90%) prevalence cancers and 57 (0.32%) incidence cancers were detected. Within categories of tumor size, the distribution of pathological stage, histology, lymph node status, and distant metastases was determined. Results: Pathological stage has a strong relationship to tumor size at baseline screening (spearman r = 0.63, p = 0.013) and repeated screening (r = 0.65, p < 0.001). In the analysis of all 76 cases, lymph node status (r = 0.59, p < 0.001) and distant metastases (r = 0.55, p = 0.005) have also a strong relationship to tumor size. The percentages of cases with no metastases (N0M0) were 100% (0/21), 89% (17/19), 62% (8/13), 83% (10/12), 50% (4/8), and 33% (1/3) for the categories 10 mm or less, 11 mm to 15 mm, 16 mm to 20 mm, 21 mm to 30 mm, 31 mm to 40 mm, and 41 mm or greater, respectively. Histology for the categories 15 mm or less was localized bronchioloalveolar carcinoma in 13 cases, adenocarcinoma with mixed subtype in 11 cases, invasive adenocarcinoma in five cases, other non-small cell carcinoma in 10 cases, and small cell carcinoma in one case. Accumulated 10-year hazard ratio of above histology was 2.4, 2.5, 2.6, 4.1, and 0.7. In multivariate analyses, pathological stage was related to only tumor size (standardized regression coefficient beta = 0.59, p < 0.001) whereas histology was related to tumor size (beta = 0.43, p < 0.001) and smoking index (beta = 0.28, p = 0.016). Conclusions: These results provide direct evidence of a stage-size relationship in long-term repeated CT screening for lung cancer. Furthermore, early detection of lung cancer of 15 mm or less in diameter leads to the detection of early-stage (N0M0) lung cancer in repeated CT screening.

No significant financial relationships to disclose.

Current trend in incidence of severe interstitial lung disease (ILD) due to gefitinib in Japan

e19075

Background: Although gefitinib is a promising agent for the treatment of advanced non-small cell lung cancer, ILD occurs in Japanese patients and sever ILD sometimes becomes fatal toxicity. ILD has been reported to be associated with preexisting ILD, smoking, and poor performance status. On the other hand, it has been reported that female gender, adenocarcinoma, non-smoker, and EGFR mutation are associated with sensitivity to gefitinib. Methods: To investigate the trend in incidence of severe ILD and patient characteristics receiving gefitinib, a total of 751 patients who were administered gefitinib in National Cancer Center Hospital East between 2002 and 2008 were retrospectively reviewed. To investigate how oncologists avoid administering gefitinib to patients with preexisting ILD, pretreatment chest CT films of all patients were also reviewed by two thoracic radiologists. Results: The patient number who received gefitinib in 2002/2003/2004/2005/2006/2007/2008 were 134/141/88/93/125/98/72, respectively. In these patients, percentages of female were 36/43/42/55/59/52/54%, adenocarcinoma 76/78/93/88/95/93/92%, non-smoker 35/30/40/53/49/48/46%, respectively. Almost all of the patient receiving gefitinib was recently restricted to adenocarcinoma. Grade 3–5 severe ILD was observed in 1.9% of all patients. However, the incidence rate of sever ILD were decreased as 3.0/2.8/2.3/1.1/0.8/1.0/1.4%, respectively. The rates of patients with preexisting ILD were also decreased as 22/14/15/5/2/3/6%, respectively. Conclusions: The correct information about efficacy and severe ILD about gefitinib influenced patient selection by oncologists, and it is reasonable to select more effective and safe patients in Japan. It is estimated that the incidence rate of severe ILD should be consequently controlled by the spread of these information.

No significant financial relationships to disclose.

Mutational analysis of EGFR and KIT in thymic epithelial tumor

18049

Background: Thymic epithelial tumor has been reported to express the epidermal growth factor receptor (EGFR), and thymic carcinoma has recently been described to have overexpression of KIT. We investigated the prevalence of EGFR and KIT mutations in patients with thymoma and thymic carcinoma to explore the potential for a targeted therapy with tyrosine kinase inhibitors. Methods: Genomic DNA was isolated from 41 paraffin-embedded tumor samples including 24 thymoma and 17 thymic carcinoma. EGFR mutations in exons 18, 19 and 21, and KIT mutations in exons 9, 11, 13 and 17, were analyzed by PCR and direct sequencing. Protein expressions of EGFR and KIT were also evaluated by immunohistochemistry. Results: We detected the EGFR mutation in 2 of 20 thymoma, but none of thymic carcinoma had mutation. All of the detected EGFR mutations were missense mutations in exon 21 (L858R and G863D, respectively). Expression of EGFR was seen in 71% of thymoma and 53% of thymic carcinoma. On mutational analysis of KIT, only one thymic carcinoma displayed a missense mutation in exon 11 (L576P). Expression of KIT was observed in 88% of thymic carcinoma and 0% of thymoma. Conclusions: Our findings indicate that a small number of patients with thymic epithelial tumor exhibit somatic mutations of EGFR or KIT although expressions of EGFR or KIT are present frequently in thymic epithelial tumor. Further investigations are warranted to determine their susceptibility to tyrosine kinase inhibitors in the treatment of thymoma and thymic carcinoma with EGFR or KIT mutations.

No significant financial relationships to disclose.

Prognostic factors for response and survival of second-line chemotherapy in patients with relapsed small cell lung cancer (SCLC)

18023

Background: Despite high response rates to initial chemotherapy, the majority of patients with SCLC experience tumor progression. Previous studies showed that both the response to initial chemotherapy and the response duration are important for predicting the efficacy of second-line chemotherapy. Therefore, relapsed SCLC is commonly classified into two groups: sensitive relapse (respond to initial chemotherapy and relapse more than 3 months after the completion of initial chemotherapy) and refractory relapse (not respond to initial chemotherapy or respond but relapse within 3 months). However, prognostic factors of the second-line chemotherapy have not been fully understood. Methods: From July 1992 to December 2003, four hundred and seventy-four patients with SCLC received chemotherapy as initial treatment, subsequently 229 patients received second-line chemotherapy (144 sensitive relapse and 85 refractory relapse) in our hospital. We analyzed the association of patients’ clinical factors with response and survival of second-line chemotherapy in sensitive relapsed patients and refractory relapsed patients, separately. For sensitive relapsed patients, analyzed clinical factors were as follows: age (<70/=70), gender (M/F), response to initial chemotherapy (CR/PR), PS at relapse (<2/=2) and the extent of disease at relapse (LD/ED). For refractory relapsed patients, analyzed clinical factors were as follows: age (<70/=70), gender (M/F), response to initial chemotherapy (responder/non-responder), PS at relapse (<2/=2) and the extent of disease at relapse (LD/ED). Results: Response to second-line chemotherapy was significantly correlated with PS in sensitive relapsed patients, however, no significant factor was detected in refractory relapsed patients. For survival, PS was the only significant prognostic factor in both sensitive and refractory relapsed patients. The median survival time was 328 days (PS<2) and 128 days (PS=2) in sensitive relapsed patients (p<0.0001), while 195 days (PS<2) and 113 days (PS=2) in refractory relapsed patients (p=0.0001), respectively. Conclusions: PS was the only significant prognostic factor for survival both in sensitive and refractory relapsed SCLC.

No significant financial relationships to disclose.

What we could know from the semiannually repeated screening with low-dose helical CT in a high-risk cohort over 10 years: Update of Anti-Lung Cancer Association project

7568

Background: There have been several randomized trials of low-dose helical CT (LDCT) screening for lung cancer. However, all trials are in progress. Therefore, before they are completed, we summarized what we could know from an update of Anti-Lung Cancer Association project, which was regarded as a longer-term study with LDCT performed at shorter intervals and with a larger number of detected cancers than any other single-armed studies. Methods: Among 2,120 participants, 1,877 (mean age 64 years, 88% male, and 84% smoker) underwent semiannually repeated screening from 1993 to 2004 (median, 3.5 years). We investigated (1) survival of patients with screening detected lung cancers, (2) presence of a stage shift (indicator of a mortality benefit), (3) appropriate duration of repeated screening, (4) identification of high-risk group by age, sex, and smoking, and (5) appropriate screening intervals by high-risk group (6 months or 1 year). Results: (1) The 5- and 10-year survival rates were 84.5% and 84.5%, respectively, in repeated screening group (n = 57) and were 68.7% and 38.1%, respectively, in initial screening group (n = 19) (P = 0.208). (2) Only in invasive adenocarcinoma, both proportion of stage II to IV and tumor size were negatively correlated with duration of repeated screening (r = −0.77, P = 0.007 and r = − 0.60, P = 0.029, respectively). (3) Detection rate of all incidence cancers were positively correlated with duration of repeated screening (r = 0.50, P = 0.020). However, detection rate remained unchanged until 5 years of repeated screening. Moreover, stage shift did not occur until 5 years of repeated screening. (4) Female sex (HR 2.9, P = 0.015) and smoking (HR 2.7, P = 0.046) were demographic risk factors for lung cancer detection at repeated screening. The accumulated 10-year detection rates for female smokers (n = 91), male smokers (n = 1,557), and non-smokers (n = 229) were 15.1%, 6.2%, and 4.3%, respectively (P = 0.002). (5) The estimated relative cancer detection powers of annual screening to semiannual screening were 50% and 57% for female and male smokers with lung cancer, respectively. Conclusions: Semiannually repeated LDCT screening over 5 years might be beneficial to smokers, especially female smokers.

No significant financial relationships to disclose.

Clinical predictive factors for mutations in epidermal growth factor receptor (EGFR) exon 19 and 21 in Japanese patients with lung cancer

18131

Background: It has been reported that lung cancer with activating mutations of EGFR are highly sensitive to gefitinib or erlotinib. Furthermore, the incidence of EGFR mutations is significantly higher in adenocarcinoma, female, never-smoker, Japanese patients. However, it is difficult to take sufficient amount of biopsy specimen of lung cancer with advanced disease for mutation analyses in clinical practice. Therefore, in this study, clinical predictive factors for EGFR mutations in Japanese patients with lung cancer were clarified. Methods: EGFR exon 19 and 21 mutation status was determined in 284 tumor samples using direct sequencing, polymerase chain reaction. To identify clinical independent predictive factors for EGFR mutation, the logistic regression analysis was performed. Results: 131 tumors of them could be analyzed for mutations in EGFR exon 19 and 21. EGFR mutations were detected in 38 tumors (29%). Median age of patients with EGFR mutations was 62 years (rage 43–84); 19 male and 19 female; 18 stage I, 15 stage II, and 5 stage III disease. Cigarette- smoking history of patients with EGFR mutations included non-smoker in 20 patients, light smoker (< 10 pack-years smoker) in 4, and heavy smoker (> 10 pack-years smoker) in 14. Multivariate analyses showed that non-smoker or light smoker (p=0.007) and adenocarcinoma (p=0.044) were the significant predictive factors contributing to EGFR mutations. Gender, serum CEA level, and clinical stage were not associated with EGFR mutations. Seventy patients treated with gefitinib and 25 PR were observed, and the overall response rate was 36%. The response rate to gefitinib in patients with and without EGFR mutations were 76% and 7%, respectively. In addition, EGFR mutations were detected in 51% (19/37) of patients with adenocarcinoma and light smoker, and the response rate to gefitinib of them was 49% (18/37). Conclusions: In Japanese patients with lung cancer, non-smoker or light smoker and adenocarcinoma were significant predictive factors for mutations in EGFR exon 19 and 21.

No significant financial relationships to disclose.

The adenocarcinoma-specific stage shift: Closely-repeated low-dose helical CT screening in a high-risk cohort for 10 years

1006

Background: Trials of CT screening for lung cancer have not yet proven the presence of a stage shift, to say nothing of a mortality benefit. On the basis of an update from Anti-Lung Cancer Association project, we investigated whether the stage shift did occur during the long-term closely-repeated low-dose helical CT screening in a high-risk cohort. Methods: A total of2,120 participants (mean age 63 yrs, 87% male, and 83% smoker) underwent low-dose helical CT screening semiannually from 1993 to 2004 (median, 3.5 years). Results: Nineteen prevalence and 57 incidence lung cancers were detected. In comparison between baseline and repeated screening, the detection rate was 0.90% vs. 0.32% (P < 0.001) and the tumor size was 24mm vs. 17mm (P = 0.018). Adenocarcinoma (74% and 63%) and stage IA (58% and 79%) were observed most commonly in both screen-groups, respectively, but their proportions showed no significant difference between groups, respectively. Regarding the survival of lung cancer patients, screening type was not a significant prognostic factor. In repeated screening,the detection rate oflung cancer except bronchioloalveolar carcinoma increased significantly depending on CT repeating times (r = 0.50, P = 0.020). Moreover, the proportion of stage II-IV and the tumor size decreased significantly only in invasive adenocarcinoma (r = -0.77, P = 0.007 and r = -0.60, P = 0.029, respectively). Conclusions: The adenocarcinoma-specific stage shift did occur in a high-risk cohort. Considering larger proportion of adenocarcinoma in this population, smokers might benefit from the long-term closely-repeated low-dose helical CT screening.

No significant financial relationships to disclose.

A phase II trial of paclitaxel plus carboplatin in patients with non-small cell lung cancer previously treated with chemotherapy

17103

Background: The purpose of this phase II trial was to evaluate the efficacy and toxicity of paclitaxel plus carboplatin in the treatment of advanced non-small cell lung cancer (NSCLC) previously treated with chemotherapy. Methods: Eligible patients were aged 20 to 75 years with a performance status (PS) 0 to 1, previous treatment with one or two prior chemotherapy regimen for advanced NSCLC, and adequate organ function. Paclitaxel 200 mg/m2 was infused over 3 hours, before carboplatin (area under the curve 6; Calvert formula) infused over 1 hour, once every 3 weeks for six cycles maximum. The primary endpoint of this study was response rate. Simon’s two-stage design was used to determine the sample size and the planned accrual was 30 patients. Results: Thirty patients were enrolled. Complete response was observed in one patient and partial response in eleven patients for an overall response rate of 40% (95% confidence interval: 23–59%). The median time to progression was 5 months. The median survival was 9.9 months (range, 2.5–33.8 months), and the 1-year survival rate was 46%. The hematologic toxicities were 7% for grade 4 neutropenia, 3% for grade 3 febrile neutropenia, 17% for grade 3 and 4 anemia, and 7% for grade 3 thrombocytopenia. Grade 3 non-hematologic toxicities included rash (7%), myalgia/arthralgia (3%), arrhythmia (3%), and infection (3%). There were no toxic deaths. Conclusions: The combination of paclitaxel plus carboplatin is an active and well-tolerated regimen in the treatment of patients with NSCLC previously treated with chemotherapy who have a good PS.

No significant financial relationships to disclose.

Trends in demographics and survival for patients (pts) with advanced non-small cell lung cancer (NSCLC)

7114

Background: In clinical trials, concurrent chemoradiotherapy improves survival of pts with inoperable stage III NSCLC compared with sequential chemoradiotherapy or radiotherapy alone, and platinum doublets with third-generation chemotherapy prolong survival of pts with stage IV NSCLC. Epidermal growth factor receptor antagonist is active as second-line chemotherapy. Few outcomes research regarding demographics and survival trends in advanced NSCLC has been conducted. Methods: The National Cancer Center Hospital East Database was searched for all pts with inoperable stage III and IV NSCLC. Data was recorded for histology, age, sex, smoking history, tumor location, stage, performance status (PS), and treatment modality. Pts were divided into two groups; group A: newly registered pts from July 1992 to December 1997, group B: from January 1998 to June 2004. Survival curves were evaluated using the Kaplan-Meier methods, and statistical significance was estimated by the log-rank test. Results: 2,135 pts (771 pts group A vs. 1364 pts group B) were identified. Pts demographics of each group (% group A/B) were as follows; male; 74/77, non-smoker; 19/19, PS 0–1; 79/85, squamous cell histology; 26/24, stage III; 45/44. Median age was 63 years old in group A and 64 in group B. Median survival (MS), 1-year survival rate (1ys), 2ys and 3ys were 8 months (M), 33%, 12% and 7% in group A, 10M, 42%, 23% and 14% in group B, respectively (P < .0001). In pts with stage III, MS, 1ys, 2ys, 3ys were 12M, 46%, 20%, 12% in group A and 15M, 56%, 34%, 20% in group B (P < .0001). In non-smoker, MS, 1 ys, 2 ys were 11 M, 46%, 19% in group A and 15 M, 56%, 30% in group B (P < .0001). In females, MS, 1 ys, 2 ys were 10 M, 39%, 16% in group A and 15 M, 55%, 31% in group B (P < .0001). Conclusion: Although there were no apparent differences in demographics between the two groups, survival was significantly improved chronologically. The improvement was prominent in stage III, non-smokers and females suggesting the benefit of chemoradiotherapy and tyrosine kinase inhibitors.

[Table: see text]