Search

e20012 Background: Smoking and lymph node metastasis (LNM) are widely known to be the important risk factor for the induction and poor prognosis of non-small cell lung cancer (NSCLC) , respectively. However, the association between smoking status and LNM in Chinese patients with NSCLC is largely unknown. Methods: The patients with pathologically diagnosed T1 NSCLC undergoing surgical resection during 2014 to 2017 were retrospectively reviewed based on LinkDoc database. Patients without smoking status in the case records were excluded. Chi-square test was used for comparison between groups. Results: Of the 10,622 patients enrolled, 1,760 patients had LNM (16.6%). For smoking status, 3,260 (30.7%) patients were recorded with smoking history, with an average smoking index (SI) of 221.2. The incidence of LNM especially N2 station LNM in patients who ever smoked was significantly higher than those of never smokers in the enrolled population (22.8% vs. 13.6%, p < 0.001; N2: 14.3% vs. 9.1%, p < 0.001). Similarly, a larger proportion of current smokers than never smokers had LNM (24.0% vs. 13.6%, p < 0.001), and also more current smokers in only N2 station LNM (14.1% vs. 9.1%, p < 0.001). However, no statistically difference was observed in the proportion of patients with LNM particularly with N2 LNM between former and current smokers (24.0% vs. 22.8%, p = 0.401; N2: 14.1% vs. 14.3%, p = 0.887). Compared with non to light smoking (SI < = 400), patients with middle to heavy smoking ( > 400) were found to be more likely to develop LNM (N1/N2: 24.8% vs. 14.6%, p < 0.001; N2: 14.3% vs. 9.8%, p < 0.001). Conclusions: These findings from the real-world study indicated a close association between smoking both former and current and LNM especially N2 station LNM in patients with T1 NSCLC. [Table: see text]

e15522 Background: This study aimed to explore the efficacy, safety of apatinib and analyze the effects of HER2 mutation status and treatment lines on gastric cancer (GC) patients (pts) treated with apatinib in real-world clinical practice. Methods: We retrospectively analyzed the study data from Linkdoc database with 270 pts who were pathologically or cytologically diagnosed GC and treated with apatinib during January 2015 to November 2018 in Henan Cancer Hospital. Survival was estimated by Kaplan-Meier method. Results: In this study, there were 180 (66.7%) male, with median age of 59 years. The vast majority of pts (259/95.9%) had adenocarcinoma and 86.3% of pts were in stage IV. HER2 was positive in 33.0% of 100 treated pts undergoing HER2 mutation testing. Apatinib was mainly used as second-line treatment (122/45.2%), followed by third-line (82/30.4%), first-line (56/20.7%), fourth- and further-line (22/8.1%), adjuvant (12/4.4%) and neoadjuvant treatment (1/0.4%). Pts received apatinib alone or combined with chemotherapy were 175 (64.8%) and 119 (44.1%), with 53.6% administered at an initial dose of 500 mg. Of all the 270 enrolled pts, the median progression free survival (mPFS) and median overall survival (mOS) were 4.3 months (95% CI: 3.5-5.0) and 6.1 months (95% CI: 5.1-8.4), respectively. For treatment lines subgroup, the mOS was 12.6 months (95% CI: 3.5-NE) in adjuvant treatment; 8.9 months (95% CI: 4.3-12.6) in first-line; 7.3 months (95% CI: 5.0-9.7) in second-line; 6.2 months (95% CI: 5.1-9.9) in third-line; 4.0 months (95% CI: 1.6-6.9) in fourth- and further-line treatment. For HER2-negative pts, the mOS was slightly longer than those of HER2-positive pts (5.7 months vs. 4.5 months), however, the difference was not statistically significant ( p = 0.5185). The most common adverse events were fatigue (25.9%), anemia (24.8%), hypertension (9.3%) and vomiting (9.3%). Conclusions: This real world study revealed that the clinical efficacy of apatinib in GC pts was satisfying and the toxicity was tolerable and controllable. Pts received apatinib in ≤ second-line treatment may gain better survival benefits, and little difference was found in survival outcomes between pts with or without HER2 mutations.

e18047 Background: Computer-aided diagnosis based on deep learning methodologies has demonstrated great potential to assist disease diagnosis with accuracy and efficiency. Specifically, the quantitative and qualitative analysis on lung nodules has proven to be important for the early-detection of lung cancer as well as its treatment in clinical practice. This study developed a 3D CNN model to facilitate the classification of pulmonary nodules. Methods: We collected 3956 lung CT scans (slice thickness≤3mm) with multiple lung nodules from 15 Class-A hospitals in China, 1155 lung CTs from Luna 16 dataset and Data Science Bowl 2017. There were 30 senior radiologists responsible for annotation and each CT scan was annotated by two of them randomly. Another 4 senior associate chief physicians were divided into two groups, each group was responsible for arbitration when conflicts occurred between the annotation doctors. All the annotated CTs were randomly selected and split to construct training, validation and test sets. We pre-processed the CTs and utilized 3D CNNs to classify these nodules as solid, partial-solid, ground glass opacity, calcified, pleural solid and pleural calcified. ROC analysis was used, and the classification capability was assessed by classification accuracy and the AUC score. Results: Table shows the overall results. The proposed model yielded an AUC score of 0.97 for the ground glass opacity and 0.90 for calcified nodules in the training set, while the AUC of them were 0.93 and 0.93 respectively in the validation set. For the test set, we got an AUC score of 0.94 for the ground glass opacity. The average classification time for each nodule was less than 0.005 sec. Conclusions: Our model may assist clinical diagnosis of lung cancer and increase its objectivity and accuracy, and the fast processing speed proves its feasibility to be applied in real clinical practice. In the future, we will enrich the dataset with clinical and genetic information, thus improving our model to boost its performance. [Table: see text]

e13022 Background: The study aimed to explore clinical correlation between venous thrombosis and driving genes status in cancer patients (pts). Methods: The clinical data of cancer pts from 2014 to 2018 in our hospital were collected to calculate the overall thrombosis rate, driving gene mutation rate, and thrombosis rate in pts with mutated driving genes. Pts were divided into mutation group and non-mutation group based on the driving gene mutation status. Khorana model was used to divide each group into three subgroups and the thrombosis rate of each subgroup was calculated. The diagnostic value of the Khorana model was calculated in the whole patient set and each group. The correlation between driving gene mutation status and thrombus status was also analyzed. Results: A total of 2049 pts were enrolled, with 4.44% overall thrombosis rate and 72.13% driving gene mutation rate. The thrombosis rates in pts with EGFR, c-Met, ALK, HER-2, KRAS or ROS1 mutations were, in turn, 5.34%, 4.33%, 3.92%, 3.73%, 3.45% and 2.74%, however, there was no statistical difference in thrombosis rates between pts with or without the above gene mutations (EGFR, p = 0.90; c-Met, p = 0.23; ALK, p = 0.85; HER-2, p = 0.28; KRAS, p = 1.00; ROS1, p = 0.58). In the correlation analysis between the driving gene status and the thrombus status, the Spearman correlation coefficient was -0.024, and no statistically significant difference was observed (p = 0.69). Based on the Khorana model, the thrombosis rates of the low-, intermediate- and high-risk subgroups in the mutation group and non- mutation group were 0% vs 0%, 10.23 vs 8.33%, and 6.67% vs 12.50% respectively, which didn’t show statistically significant differences (low- risk, p = 1.00; intermediate-risk, p = 0.85; high-risk, p = 0.59). For the whole patient set, mutation group and non-mutation group, the Khorana model predicted venous thrombosis with the sensitivities of 13.33%, 10.00% and 20.00%, as well as the specificities of 92.02%, 92.55% and 90.67%; and the area under the ROC curve were 0.706 (p = 0.007), 0.719 (p = 0.02) and 0.681 (p = 0.18), respectively. Conclusions: The Khorana model has a moderate diagnostic value, and there is no correlation between the driving gene status of cancer pts and venous thrombosis.

e14686 Background: According to previous reports, 1% to 30% patients with malignant tumors developed venous thromboembolism (VTE) and an increased risk of VTE in hospitalized patients. However, real-world clinical analysis of VTE in Chinese hospitalized patients with malignant solid tumors is scarce. Methods: We utilized the collaborative Linkdoc database to retrospectively analyze hospitalized cancer patients in Department of Medical Oncology of the Fourth Hospital of Hebei Medical University from May 2014 to November 2018. Patients who were pathologically confirmed malignant solid tumor, ≥ 18 years of age, ECOG PS 0-4, median overall survival ≥ 1 months, and absence of serious diseases of vital organs such as heart, liver, kidney and blood system or the above chronic diseases in inactive period were included in this study. Results: A total of 3831 patients were enrolled, of whom 180 (4.70%) patients had VTE. The highest incidence of VTE in solid tumors was gynecological cancer (61/1288, 11.90%), followed by gastric cancer (42/701, 5.71%), other cancers (11/206, 5.33%), lung cancer (61/1288, 4.74%), colorectal cancer (18/399, 4.51%), breast cancer (34/903, 3.77%), and esophageal cancer (6/250, 2.40%). Of the enrolled patients, 21.11% (809/3831) underwent relevant imaging examinations to screen for VTE as required by the guidelines of cancer associated VTE. Based on the imaging examinations, the anticoagulant rates of patients with or without VTE were 100% (61/61) and 34.09% (255/748), respectively. Notably, only one patient received the anticoagulant therapy (both anticoagulant dose and time) in lines with the guidelines. Conclusions: This single-center real-world study showed the overall rates of VTE in hospitalized cancer patients was similar with the data previously reported. Referring to the guidelines of cancer associated VTE, the proportion of Chinese cancer patients receiving normative imaging examinations and anticoagulant therapy is seriously low currently, which should be paid more attention to in clinical practices by Chinese medical workers.

e13130 Background: Next-generation sequencing (NGS) examination based on liquid biopsy has been agreed as a kind of routine diagnostic tool in cancer field and may provide new sights about oncogenesis mechanisms. This study aimed to observe the genomic alteration features of common cancers and their correlation with clinical characters. Methods: We enrolled 39 subjects (38 subjects with cancer and 1 healthy subject) and administrated mutation examination using a panel including 38 genes, which targeted 6 cancer types. For each subject, the genomic DNA was extracted and NGS was performed. Results: All patients were found to carry at least one alteration among the 38 genes. PSM2 showed a highest mutation rate, with 77% subjects carrying mutations in this gene. Besides, MSH2 and STK11, etc. exhibited a high mutation rate. Overall 17 genes were found to contain known pathogenic or outcome-unknown mutations, among which 15 had outcome-unknown mutations and 5 had recognized pathogenic mutations. RAD50 had a highest frequency of outcome-unknown mutations, and secondly MLH1 and MRE11A. The clinical characters including sex, metastasis/recurrence, and family cancer history had relationships with mutation types. Conclusions: Genomic DNA sequencing is a feasible and minimally invasive approach in cancer genetic analysis and a promising tool in prediction of cancer onset and prognosis. A large amount of variations are associated with sex, family history and cancer types, and may decide metastasis/recurrence outcomes. Those who have a family history of cancer are recommended to receive NGS examination.

e20016 Background: Accurate lymph node assessment is crucial for clinical patient management and prognostic prediction. However, the procedures may not always be performed preoperatively due to time, cost or risk of procedure itself. Present study aimed to investigate the real world clinical practice of preoperative N staging in surgically resected pathologic T1 lung cancer patients in China. Methods: This is a multicenter retrospective study. Real world data on clinical and pathologic N staging was retrieved from electronic medical records in 10 thoracic surgery centers (LinkDoc Database). Kappa index was calculated to determine the agreement between clinical and pathologic N stage. Results: A total of 11360 patients who underwent surgical resection for pT1 lung cancer were identified during January 2014 to September 2017. Clinical N staging was available for 6057 (53.3%) patients. Of which, 959 (15.8%) patients were achieved by invasive procedures. Considering the 5795 patients with clinical and pathologic N staging data, the overall agreement was 89.0% with a kappa index of 0.4220 (95%CI: 0.3842-0.4598). This was 96.4% for N0, 22.1% for N1 and 39.8% for N2 when considering pathologic N stage as gold standard (Table). In general, upstaging (7.3%) was more common in our population. Conclusions: Over 50% of pathologic T1 patients had preoperative N staging in our clinical practice. Clinical N staging showed low accuracy for predicting lymph node status. The findings of our study highlight the increasing attention on clinical N staging and procedure optimization in China. [Table: see text]

e20523Background: Clinical characteristics of lymph metastases are hetergeneous in the patients with T1 non-small cell lung cancer (NSCLC). Preoperative assessment of lymph metastases is essential ...

e20527Background: Radical resection is preferred for treatment of patients with T1 lung cancer, and patient prognoses could be significantly affected by tumor pathological types. Since the clinical...

e20534Background: Even though anatomical lobectomy with dissection of systematic mediastinal lymph nodes has been the standard treatment for the patients with non-small cell lung cancer (NSCLC), th...

e20522Background: The golden surgery standard of early-stage non-small cell lung cancer (NSCLC) is lobectomy with dissection of mediastinal lymph nodes. Due to the risks of reduced pulmonary reserv...

e20525Background: Lung cancer is the leading cause of cancer related deaths in China. Along with rapid development of new technologies, the modes of surgeries are experiencing changes. The purposes...

e20535Background: Postoperative or perioperative systematic chemotherapies (CTs) in patients with non-small cell lung cancer (NSCLC) have been widely accepted due to providing significant clinical ...

4052Background: Ivor Lewis (Iv) and McKeown (Mc) are two commonly used minimally invasive esophagectomy. Currently, there are limited data to compare effectiveness and safety between Iv and Mc in C...