Image 1.

Accurate prediction of drug responses in cancer cell lines (CCLs) and transferable prediction of clinical drug responses using CCLs are two major tasks in personalized medicine. Despite the rapid advancements in existing computational methods for preclinical and clinical cancer drug response (CDR) prediction, challenges remain regarding the generalization of new drugs that are unseen in the training set. Herein, we propose a multimodal fusion deep learning (DL) model called drug-target and single-cell language based CDR (DTLCDR) to predict preclinical and clinical CDRs. The model integrates chemical descriptors, molecular graph representations, predicted protein target profiles of drugs, and cell line expression profiles with general knowledge from single cells. Among these features, a well-trained drug-target interaction (DTI) prediction model is used to generate target profiles of drugs, and a pretrained single-cell language model is integrated to provide general genomic knowledge. Comparison experiments on the cell line drug sensitivity dataset demonstrated that DTLCDR exhibited improved generalizability and robustness in predicting unseen drugs compared with previous state-of-the-art baseline methods. Further ablation studies verified the effectiveness of each component of our model, highlighting the significant contribution of target information to generalizability. Subsequently, the ability of DTLCDR to predict novel molecules was validated through in vitro cell experiments, demonstrating its potential for real-world applications. Moreover, DTLCDR was transferred to the clinical datasets, demonstrating satisfactory performance in the clinical data, regardless of whether the drugs were included in the cell line dataset. Overall, our results suggest that the DTLCDR is a promising tool for personalized drug discovery.

Accurate prediction of drug responses in cancer cell lines(CCLs)and transferable prediction of clinical drug responses using CCLs are two major tasks in personalized medicine.Despite the rapid advancements in existing computational methods for preclinical and clinical cancer drug response(CDR)prediction,chal-lenges remain regarding the generalization of new drugs that are unseen in the training set.Herein,we propose a multimodal fusion deep learning(DL)model called drug-target and single-cell language based CDR(DTLCDR)to predict preclinical and clinical CDRs.The model integrates chemical descriptors,mo-lecular graph representations,predicted protein target profiles of drugs,and cell line expression profiles with general knowledge from single cells.Among these features,a well-trained drug-target interaction(DTI)prediction model is used to generate target profiles of drugs,and a pretrained single-cell language model is integrated to provide general genomic knowledge.Comparison experiments on the cell line drug sensitivity dataset demonstrated that DTLCDR exhibited improved generalizability and robustness in predicting unseen drugs compared with previous state-of-the-art baseline methods.Further ablation studies verified the effectiveness of each component of our model,highlighting the significant contribution of target information to generalizability.Subsequently,the ability of DTLCDR to predict novel molecules was validated through in vitro cell experiments,demonstrating its potential for real-world applications.Moreover,DTLCDR was transferred to the clinical datasets,demonstrating satisfactory performance in the clinical data,regardless of whether the drugs were included in the cell line dataset.Overall,our results suggest that the DTLCDR is a promising tool for personalized drug discovery.

Objective:To investigate the predictive ability of Glypican-3 (GPC3) positive hepatocellular carcinoma based on the hepatobiliary specific contrast agent gadoxetate disodium enhancement of the liver imaging reporting and data system version 2018 (LI-RADS v2018) imaging features, and to assess the relevant clinical imaging features for postoperative recurrence in GPC3 positive HCC patients.Methods:This study was a cohort study. A total of 122 hepatocellular carcinoma patients who underwent gadoxetate disodium enhanced MRI examination with hepatic tumor resection in Henan Provincial People′s Hospital from January 2017 to December 2021 were retrospectively collected, including 96 GPC3 positive and 26 GPC3 negative patients. The imaging features defined by LI-RADS v2018 of HCC lesions were analyzed. Patients were followed up for 40 months to determine recurrence free survival (RFS). The logistic regression was used to analyze the risk factors of GPC3 positivity. An imaging model, and a clinical-imaging model which combined the patient′s alpha-fetoprotein levels were constructed. The efficacy of the model for predicting GPC3 positivity was assessed using receiver operating characteristic curves. Kaplan-Meier method was used to draw the survival curve, and the log-rank test was used to compare the RFS between GPC3 positive and negative patients. Risk factors affecting the recurrence of GPC3 positive HCC were assessed by Cox regression.Results:The results of logistic multivariate regression analysis confirmed that rim enhancement ( OR=5.685, 95% CI 1.229-26.287, P=0.026) and irregular tumor margin at hepatobiliary phase ( OR=4.431, 95% CI 1.684-11.663, P=0.003) were independent risk factors for GPC3 positive HCC. The area under the curve for predicting GPC3 positivity was 0.745 (95% CI 0.636-0.854) for the imaging model and 0.776 (95% CI 0.677-0.876) for the clinical-imaging model. The mean RFS in the GPC3 positive group was 22 months, and it was 32 months in the negative group. There was a statistically significant difference in RFS between the two groups ( χ2=5.15, P=0.023). The multivariate Cox regression analysis showed that the arterial rim enhancement ( HR=5.460, 95% CI 1.966-15.162, P=0.001), microvascular invasion ( HR=2.402, 95% CI 1.210-4.769, P=0.012), portal vein tumor thrombus ( HR=3.226, 95% CI 1.114-9.344, P=0.031) were independent risk factors for recurrence after hepatic tumor resection for GPC3-positive HCC. Conclusions:A model based on the LI-RADS v2018 imaging features of hepatobiliary specific contrast agent gadoxetate disodium enhancement can effectively predict GPC3 positive HCC. The arterial rim enhancement, microvascular invasion and portal vein tumor thrombus are independent risk factors for postoperative recurrence of GPC3 positive HCC.

Objective:To analyze the current status of the global burden of stroke disease from 1990 to 2019,to predict the development trend of stroke disease burden in the 30 years from 2020 to 2049,and to provide a basis for formulating national health policies on stroke diseases.Methods:The Global Burden of Disease Study 2019(GBD 2019)database was searched to extract global stroke disease incidence,prevalence,case fatality,and disability-adjusted life years(DALYs)disease burden indicators from 1990-2019,the trends over time were modeled using linear,Poisson,and exponential regressions,prediction and study of the relationship between stroke and sociodemographic index(SDI)based on per capita gross domestic product(GDP)were conducted.Results:The global burden of stroke disease increased significantly from 1990-2019 and is predicted to continue to rise over the next 30 years(2020-2049).In 2049,the global stroke incidence,prevalence,case fatality,and DALYs will increase by 8.53 million(63％),119.83 million(109％),7.79 million(118％)and 118.92 million person-years(79％),respectively,compared with 2019,with a significant increase in the burden of stroke in the elderly population.In the next 30 years,the age-standardized incidence rates of stroke in men and women will be similar,while the age-standardized rates of prevalence in women will be relatively higher,and age-standardized case fatality rates and DALYs in men will be relatively higher.The disease burden of stroke was negatively correlated with SDI.The burden of stroke disease was significantly higher in regions with a low SDI than in regions with a high SDI.Conclusion:The global burden of stroke will increase in the next 30 years,which may be related to the aging of population and closely related to the development of economy.It is necessary to strengthen the prevention of stroke and formulate targeted strategies targeted strategies according to different SDI regions.

Objective:To evaluate the value of quantitative analysis of the relative signal intensity (SI) of liver gadolinium disodium enhanced MRI in the grading of liver fibrosis.Methods:From January 2018 to October 2020, the relevant data of 131 patients who underwent gadoxetate disodium enhanced MRI examination were retrospectively analyzed in Henan Provincial People′s Hospital. All patients had histopathological results. According to the Laennec grading system of liver fibrosis, the patients were classified in F0-F1 (27 cases), F2 (19 cases), F3 (17 cases) and F4 (68 cases). The signal intensity of the liver, erector spinae and spleen were measured before and after the enhancement; and 5 post-contrast relative SI parameters were calculated, including the relative enhancement (RE), liver-to-muscle contrast ratio (LMC), liver-to-spleen contrast ratio (LSC), LMC increase rate, LSC increase rate. The differences of 5 post-contrast relative SI parameters among the different fibrosis grades were compared using one-way analysis of variance. The receiver operating characteristic (ROC) curves were drawn to evaluate the diagnostic efficacy of 5 post-contrast relative SI parameters in the diagnosis of clinically significant liver fibrosis (F2-F4), advanced liver fibrosis (F3-F4) and liver cirrhosis (F4).Results:The differences of RE, LMC, LSC, LMC increase rate, LSC increase rate among different liver fibrosis grades were statistically significant (all P<0.001). With the increasing of the degree of liver fibrosis, the RE, LMC increase rate and LSC increase rate showed decreased. ROC results showed that the area under the curve (AUC) of RE, LMC increase rate, LSC increase rate in diagnosing liver fibrosis in all levels were greater than those of LMC and LSC. The AUC values of RE, LMC increase rate, LSC increase rate in the diagnosis of significant fibrosis (F2-F4) were 0.89, 0.86, 0.83, with the sensitivity as 81.7%, 71.2%, 81.7%, and the specificity as 96.3%, 85.2%, and 74.1%, respectively. The AUC values of RE, LMC increase rate, LSC increase rate in the diagnosis of advanced liver fibrosis (F3-F4) were 0.93, 0.88, 0.86, with the sensitivity as 84.7%, 72.9%, 91.8%, and the specificity as 91.3%, 87.0 %, 71.7%; and the AUC values for diagnosing liver cirrhosis (F4) were 0.92, 0.86, 0.85, with the sensitivity as 82.4%, 76.5%, 92.7%, and the specificity as 88.9%, 81.0%, 65.1%, respectively. Conclusion:Gadoxetate disodium enhanced MRI relative SI parameters including RE, LMC increase rate and LSC increase rate might be used as a useful imaging marker in liver fibrosis grading.

Objective:To evaluate the outcome of endoscopic submucosal dissection (ESD) of colorectal tumors, and to analyze the factors affecting the therapeutic efficacy of ESD.Methods:Clinical data of patients with colorectal tumors who were treated with ESD in Department of Gastroenterology in Beijing Hospital from November 2016 to December 2019 were reviewed. A total of 82 patients with pathologically confirmed colorectal adenoma or carcinoma of diameter ≥20 mm were included. The clinical features, ESD and pathological outcomes of the patients were analyzed.Results:All 82 lesions were single, with the mean diameter of 29.72±10.74 mm. Lesions were divided into the laterally spreading tumors (LST), mainly located in the ascending colon and ileocecal region (47.8%, 22 / 46) and the protruding colorectal tumors, mainly located in the left colon, 52.8% (19 / 36) of which were located in the sigmoid colon. The overall resection rate was 81.7%(67/82) and the curative resection rate was 72.0%(59/82). The incidences of bleeding and perforation were 2.4%(2/82) and 1.2%(1/82), respectively. The curative resection rates [91.4%(32/35), 63.6%(7/11) and 55.6%(20/36), P=0.003] and surgical operation rates [8.6%(3/35), 18.2%(2/11) and 36.1%(13/36), P=0.010] of LST-G, LST-NG and protruding colorectal tumors were significantly different. Multivariate regression analysis showed that protruding colorectal tumor ( OR=3.396, 95% CI: 1.014-11.374, P=0.047) and submucosal severe fibrosis (F2 type) ( OR=5.508, 95% CI: 2.216-13.692, P=0.001) were independent risk factors for non-curative ESD resection of colorectal tumors. Conclusion:ESD is effective and safe for colorectal tumors of diameter ≥ 20 mm. However, there are some differences in the rate of submucosal invasion and treatment outcome among different types of lesions. The risk factors for non-curative resection are protruding tumors and severe submucosal fibrosis.

Colorectal cancer is one of the common malignancies of the digestive tract, and endoscopic submucosal dissection (ESD) has been widely used in the treatment of early colorectal neoplasms. The risk factors of non-curative resection in ESD for early colorectal neoplasms have been confirmed in many studies. This article reviewed the advances in study on risk factors of non-curative resection in ESD for early colorectal neoplasms.

OBJECTIVETo study the effect of neoadjuvant chemotherapy on nutritional status in patients with locally advanced gastric cancer.

METHODSCases inclusion criteria: (1)18-65 years old; (2) Gastric cancer confirmed by gastroscopic biopsy; (3) Preoperative TNM stage III( according to the AJCC stage 2000 standard; (4) Kamosfsky functional status score> 60 points; (5)Receiving neoadjuvant chemotherapy voluntarily and signing the informed consent form. Case exclusion criteria: (1)Having contraindications of chemotherapy and surgery; (2) Suffering from heart, liver and kidney and other underlying diseases; (3) Concurrent with malignant diseases, wasting disease or other digestive diseases. According to the above criteria, clinical data of 73 patients of stage III( gastric cancer receiving neoadjuvant chemotherapy at Weifang People's Hospital from May 2015 to March 2017 were prospectively collected. The cohort study was adopted. After removing 3 patients who did not complete the chemotherapy, a total of 70 patients who completed the chemotherapy were included in the study. All the patients received SOX chemotherapy without nutritional support during chemotherapy. Changes of body composition and nutritional indicators were analyzed before and after chemotherapy, and according to the tumor regression after chemotherapy, patients were divided into response group (complete or sub-total tumor regression) and non-response group (tumor part, with or without a small amount of retreat) for stratified analysis.

RESULTSOf 70 gastric cancer patients, 40 were male and 30 were female with a age of (53.8±6.4) (28 to 64) years. There were 26 cases (37.1%) of stage III(a, 35 cases (50.0%) of stage III(b and 9 cases (12.9%) of stage III(c. There were 41 cases in response group and 29 cases in non-response group. Three patients (4.3%) were complete remission (CR) and 38 patients (54.3%) were partial remission (PR) in response group, while 23 cases (32.9%) were stable disease (SD) and 6 cases (8.6%) were progressive disease (PD). After neoadjuvant chemotherapy, the extracellular fluid of the whole patients increased from (13.3±1.7) L to (13.5±1.6) L (t=-2.044, P=0.045); the intracellular fluid decreased from (21.4±2.5) L to (21.1±2.4) L (t=2.369, P=0.021); the lymphocyte count decreased from (0.31±0.10)×10/L to (0.29±0.10)×10/L (t=1.706, P=0.009); the other indexes were not significantly different (all P>0.05). Stratified analysis showed that after neoadjuvant chemotherapy in response group, body mass increased from (60.1±8.8) kg to (61.0±8.3) kg (t=-2.773, P=0.008); body mass index increased from (21.9±2.4) kg/m to (22.3±1.9) kg/m(t=-3.479, P=0.001), while above two parameters did not differ significantly in non-response group. No significant differences in body mass, body mass index, skeletal muscle, inorganic salt, extracellular fluid, body fat, protein, upper arm circumference and intracellular fluid were found between two groups before and after neoadjuvant chemotherapy(all P>0.05). Except slight decrease in hemoglobin and lymphocyte, the other nutritional indicators were slightly elevated in response group, while the differences were not statistically significant(all P>0.05). All nutritional indicators in non-response group were slightly decreased without significant differences as well (all P>0.05).

CONCLUSIONNeoadjuvant chemotherapy does not aggravate malnutrition in patients with locally advanced gastric cancer.

Adult ; Antineoplastic Combined Chemotherapy Protocols ; therapeutic use ; Chemotherapy, Adjuvant ; Cohort Studies ; Female ; Humans ; Male ; Middle Aged ; Neoadjuvant Therapy ; Neoplasm Staging ; Nutritional Status ; Remission Induction ; Stomach Neoplasms ; drug therapy