Background: Accurate glioma grading before surgery is of the utmost importance in treatment planning and prognosis prediction. But previous studies on magnetic resonance imaging (MRI) images were not effective enough. According to the remarkable performance of convolutional neural network (CNN) in medical domain, we hypothesized that a deep learning algorithm can achieve high accuracy in distinguishing the World Health Organization (WHO) low grade and high grade gliomas.Methods: One hundred and thirteen glioma patients were retrospectively included. Tumor images were segmented with a rectangular region of interest (ROI), which contained about 80% of the tumor. Then, 20% data were randomly selected and leaved out at patient-level as test dataset. AlexNet and GoogLeNet were both trained from scratch and fine-tuned from models that pre-trained on the large scale natural image database, ImageNet, to magnetic resonance images. The classification task was evaluated with five-fold cross-validation (CV) on patient-level split.Results: The performance measures, including validation accuracy, test accuracy and test area under curve (AUC), averaged from five-fold CV of GoogLeNet which trained from scratch were 0.867, 0.909, and 0.939, respectively. With transfer learning and fine-tuning, better performances were obtained for both AlexNet and GoogLeNet, especially for AlexNet. Meanwhile, GoogLeNet performed better than AlexNet no matter trained from scratch or learned from pre-trained model.Conclusion: In conclusion, we demonstrated that the application of CNN, especially trained with transfer learning and fine-tuning, to preoperative glioma grading improves the performance, compared with either the performance of traditional machine learning method based on hand-crafted features, or even the CNNs trained from scratch.

Lung adenocarcinoma (LUAD) is a common type of lung cancer with high frequent metastasis and a high death rate. However, genes responsible for LUAD metastasis are still largely unknown. Here, we identify an important role of ras homolog family member V (RHOV) in LUAD metastasis using a combination of bioinformatic analysis and functional experiments. Bioinformatic analysis shows five hub LUAD metastasis driver genes (RHOV, ZIC5, CYP4B1, GPR18 and TCP10L2), among which RHOV is the most significant gene associated with LUAD metastasis. High RHOV expression predicted shorter overall survival in LUAD patients. RHOV overexpression promotes proliferation, migration, and invasion of LUAD cells, whereas RHOV knockdown inhibits these biological behaviors. Moreover, knockdown of RHOV suppresses LUAD tumor growth and metastasis in nude mice. Mechanistically, RHOV activates Jun N-terminal Kinase (JNK)/c-Jun signalling pathway, an important pathway in lung cancer development and progression, and regulates the expression of markers of epithelial-to-mesenchymal transition, a process involved in cancer cell migration, invasion and metastasis. RHOV-induced malignant biological behaviors are inhibited by pyrazolanthrone, a JNK inhibitor. Our findings indicate a critical role of RHOV in LUAD metastasis and may provide a biomarker for prognostic prediction and a target for LUAD therapy.

Conclusions:The dynamic change of CEA, CA125, CYFRA21-1, and SCC-Ag from baseline have prognostic value for late-stage NSCLC patients treated with PD-1/PD-L1 inhibitors. Decrease of associated biomarkers serum levels were associated with favorable clinical outcomes.

Background: Immune checkpoint inhibitors (ICIs) represent a great breakthrough in the treatment of advanced non-small cell lung cancer (aNSCLC). However, whether immunotherapy beyond progression (IBP) is effective for aNSCLC has yet to be established. Therefore, a retrospective clinical study was conducted to investigate the efficacy of IBP in patients with aNSCLC under real-world conditions.Methods: A total of 125 Chinese patients with aNSCLC who experienced progressive disease (PD) after receiving monotherapy or combination therapy (combined with chemotherapy or/and antiangiogenic therapy) with programmed cell death-1 (PD-1)/programmed cell death ligand-1 (PD-L1) inhibitors between January 2015 and March 2019 were enrolled. Patients who were treated with ICIs for more than 6 weeks after PD were defined as IBP (n=39), while those who received ICI treatment for less than 6 weeks or discontinued it due to PD were defined as non-IBP (n=86). Patient clinical characteristics were evaluated.An optimization-based method was applied to balance the clinical baseline characteristics between the two groups.Results: In total population, the IBP group had longer overall survival (median OS, 26.6 vs. 9.5 months; HR, 0.40; 95% CI: 0.23-0.69; P<0.001) and progression-free survival (median PFS, 8.9 vs. 4.1 months; HR, 0.41; 95% CI: 0.26-0.65; P<0.001), compared with the non-IBP group. Despite no significant difference in objective response rate (ORR, 15.4% vs. 11.6%, P=0.560), disease control rate (DCR) was significantly higher in the IBP group (89.7% vs. 61.6%, P<0.001). After balancing baseline covariates, the IBP group also had longer OS (median: 26.6 vs. 10.7 months; HR, 0.40; 95% CI: 0.19-0.84; P=0.015) and PFS (median: 9.7 vs. 4.3 months; HR, 0.28; 95% CI: 0.15-0.51; P<0.001), with a benefit in either of patients previously treated with ICI monotherapy or in combination therapy and with non-response to the previously ICI.Conclusions: IBP is associated with longer OS and PFS in patients with aNSCLC. Our findings may suggest new therapeutic options for patients with aNSCLC who experienced disease progression after initial immunotherapy.

Background: Targeting immune checkpoints represents an immense breakthrough in cancer therapeutics. The prognostic value of hemoglobin (Hb) has been investigated in many malignancies including non-small cell lung cancer (NSCLC). However, the prognostic impact of pretreatment Hb count for immune checkpoint inhibitors (ICIs) in advanced NSCLC patients remains unclear. Methods: A total of 310 late-stage NSCLC patients who received ICI therapies between January 2015 and March 2019 were prospectively enrolled. We used a propensity score-matched cohort analysis for this study. Patients’ clinicopathological characteristics and pretreatment Hb concentration were assessed against the progression-free survival (PFS) and overall survival (OS) using the Kaplan–Meier method and Cox proportional hazards regression. Results: A propensity score (PS)-matched cohort analysis was applied to adjust for potential bias and to create two comparable groups according to patients’ clinicopathological characteristics. The patients with normal baseline Hb levels (⩾110 g/L) had significantly longer PFS [median: 10.0 versus 4.0 months, hazard ratio (HR): 0.63, 95% confidence interval (CI): 0.46−0.86; p = 0.001] and OS [median: 17.6 versus 10.5 months, HR (95% CI): 0.56 (0.40−0.79); p < 0.001] than those with decreased Hb count (<110 g/L) in a PS-matched cohort ( n = 255). For patients with normal pretreatment Hb levels, ICI combination therapy was significantly associated with better PFS [median: 11.1 versus 8.0 months, HR (95% CI): 0.74 (0.50−1.06); p = 0.09] and OS [median: 26.0 versus 12.9 months, HR (95% CI): 0.56 (0.37−0.86); p = 0.008] than monotherapy, but there was no such trend for patients with decreased baseline Hb levels. Conclusion: Our findings showed that normal pretreatment Hb count served as a favorable prognostic marker in advanced NSCLC patients treated with ICIs, representing an economical biomarker with readily measuring performance among all reported ones.