Early prediction of mortality risk among patients with severe COVID-19, using machine learning.

BACKGROUND: Coronavirus disease 2019 (COVID-19), caused by severe acute respiratory syndrome coronavirus 2 infection, has been spreading globally. We aimed to develop a clinical model to predict the outcome of patients with severe COVID-19 infection early. METHODS: Demographic, clinical and first laboratory findings after admission of 183 patients with severe COVID-19 infection (115 survivors and 68 non-survivors from the Sino-French New City Branch of Tongji Hospital, Wuhan) were used to develop the predictive models. Machine learning approaches were used to select the features and predict the patients' outcomes. The area under the receiver operating characteristic curve (AUROC) was applied to compare the models' performance. A total of 64 with severe COVID-19 infection from the Optical Valley Branch of Tongji Hospital, Wuhan, were used to externally validate the final predictive model. RESULTS: The baseline characteristics and laboratory tests were significantly different between the survivors and non-survivors. Four variables (age, high-sensitivity C-reactive protein level, lymphocyte count and d-dimer level) were selected by all five models. Given the similar performance among the models, the logistic regression model was selected as the final predictive model because of its simplicity and interpretability. The AUROCs of the external validation sets were 0.881. The sensitivity and specificity were 0.839 and 0.794 for the validation set, when using a probability of death of 50% as the cutoff. Risk score based on the selected variables can be used to assess the mortality risk. The predictive model is available at [https://phenomics.fudan.edu.cn/risk_scores/]. CONCLUSIONS: Age, high-sensitivity C-reactive protein level, lymphocyte count and d-dimer level of COVID-19 patients at admission are informative for the patients' outcomes.

Low expression of RECK in oral squamous cell carcinoma patients induces a shorter survival rate through an imbalance of RECK/MMPs.

BACKGROUND: RECK, as a negative MMP regulator, is extensively expressed in normal cells but decreased in tumors. In OSCC, the relationship between RECK and MMPs and the potential prognostic impact remains unclear. In this research, for the first time, we investigated the expression of RECK associated with MMPs during OSCC carcinogenesis in a large sample and its association with 5-year survival rate. MATERIAL AND METHODS: Immunohistochemical SP technique was applied to study the expression of RECK and MMP-2 and MMP-9 in 108 cases of OSCC and 30 normal oral mucosae. Univariate and multivariate Cox regression analysis was utilized for disease-free survival and overall survival, and analyzed by Kaplan-Meier method regarding RECK expression in patients of OSCC. RESULTS: We found lower expression of RECK in OSCC was 51.85% (56/108) compared with 93.33% (28/30) in the control group. However, the higher expression of MMP-2 and MMP-9 was 74.07% (80/108) and 70.37% (76/108) in OSCC, respectively, compared with 20% (6/30) and 13.3% (4/30) in the control group. Furthermore, the decrease of RECK expression and the increase of MMP-2, and -9 expression were significantly correlated with the loss of histologic differentiation, the occurrence of lymphatic metastasis, and the increase of OSCC clinical stage (P<0.05). OSCC patients with a low level of RECK expression had a lower rate of 5-year survival. CONCLUSION: RECK may prevent metastasis and improve OSCC patients' prognosis through a RECK/MMP-2, and -9 imbalance. Furthermore, RECK is a prospective prognostic indicator and therapeutic target for cancer molecular targeting therapy. Low expression of RECK may be a significant negative prognostic predictor.

Nomograms for predicting long-term overall survival and cancer-specific survival in lip squamous cell carcinoma: A population-based study.

BACKGROUND: The goal of this study was to establish and validate two nomograms for predicting the long-term overall survival (OS) and cancer-specific survival (CSS) in lip squamous cell carcinoma (LSCC). METHODS: This study selected 4175 patients who were diagnosed with LSCC between 2004 and 2015 in the SEER (Surveillance, Epidemiology, and End Results) database. The patients were allocated randomly to a training cohort and validation cohort. Variables were selected using a backward stepwise method in a Cox regression model. Based on the predictive model with the identified prognostic factors, nomograms were established to predict the 3-, 5-, and 8-year survival OS and CSS rates of LSCC patients. The accuracy of the nomograms was evaluated based on the consistency index (C-index), while their prediction accuracy was evaluated using calibration plots. Decision curve analyses (DCAs) were used to evaluate the performance of our survival model. RESULTS: The multivariate analyses demonstrated that age at diagnosis, marital status, sex, race, American Joint Committee on Cancer stage, surgery status, and radiotherapy status were risk factors for both OS and CSS. The C-index, area under the time-dependent receiver operating characteristic curve, and calibration plots demonstrated the good performance of the nomograms. DCAs of both nomograms further showed that they exhibited good 3-, 5-, and 8-year net benefits. CONCLUSIONS: We have developed and validated LSCC prognosis nomograms for OS and CSS for the first time. These nomograms can be valuable tools for clinical practice when clinicians are helping patients to understand their survival risk for the next 3, 5, and 8 years.

[Construction of antisense human tankyrase-1 RNA retroviral vector and its inhibition on tongue cancer cells].

OBJECTIVE: To investigate the inhibition of telomerase activity and cellular proliferation in tongue cancer TCCA-8113 cell lines by antisense human tankyrase-1 RNA treatment, and explore the possibility of the tankyrase-1 as a target of gene therapy for tongue cancer. METHODS: The replication deficient retrovirus expressing tankyrase-1 antisense RNA was constructed to infect the TCCA-8113 cells. Tankyrase-1 expression was examined by RT-PCR. Telomerase activity was assayed by telomerase repeat amplification protocol (TRAP). Cell proliferation was investigated by cellular growth curve. Cellular apoptosis was detected by flow cytometry method and invert microscope. RESULTS: Tankyrase-1 expression and telomerase activity of tongue cancer TCCA-8113 cells were significantly inhibited. There was G(1)-S phase arrest when TCCA-8113 cells were treated with antisense tankyrase-1 transduction. Cellular proliferation was arrested, and cellular apoptosis occurred after antisense tankyrase transduction. CONCLUSIONS: The transduction of antisense tankyrase-1 by retroviral vector can significantly inhibit the tankyrase-1 expression and telomerase activity of tongue cancer TCCA-8113 cell lines, and arrest the cellular proliferation and promote cellular apoptosis. The tankyrase may be a potential target of gene therapy for tongue cancer.

Management of mandibular hypoplasia using distraction osteogenesis technique.

By using distraction osteogenesis technique, 3 cases of mandibular hypoplasia were treated by home-made and German-made jaw distractors: including one patient suffered from bilateral ankylosis of temporo-mandibular joint and 2 patients from deficiency of mandible. The duration of distraction osteogenesis was one month. The bone distractor was removed 3 months after operation. Satisfactory results were obtained in all 3 cases. Distraction osteogenesis can successfully be used in mandibular functional reconstruction and has much more advantages than traditional technique.

[The study of magnetic resonance sialography in diagnosing parotid diseases].

OBJECTIVE: To study the feasibility of diagnosing parotid disease with magnetic resonance sialography (MRS) and to select the optimal scanning sequence. METHODS: Twenty-three patients with parotid gland disease and 5 normal adults received sialography using magnetic resonance imaging technique and several sequences (including IR-FSE, FSE, SS-IR-FSE, SS-FSE) were used. After first scanning, the patients were scanned respectively 3 and 10 minutes after buccal application of vitamin C. And MR images of duct obtained. The images of parotid duct system were analysed and evaluated according to their displaying effects. Qualitative diagnosis was made based on MRI and those diagnosis were compared with pathological diagnosis after operation. RESULTS: Images of MR sialography clearly displayed the main duct and its branches of parotid gland and the pathological changes of duct, such as narrow, expanded, stoppage. Of the scanning sequences, IR-FSE was superior to others in manifesting the parotid gland duct (P < 0.05). The performance of images after being given vitamin C did not significantly improve the displaying effect. The accurate rate of qualitative diagnosis was 95.6%. CONCLUSIONS: MR sialography can clearly display the parotid ducts and their pathological changes. The accurate rate of qualitative diagnosis of parotid disease was higher than that X-ray sialography.

The application of virtual endoscopy with computed tomography in maxillofacial surgery.

OBJECTIVE: To use virtual endoscopy to obtain detailed three-dimensional images of bone, soft tissue and paranasal sinus damage as well as images of adjacent unaffected areas for optimum surgical planning. METHODS: A spiral CT scanner was used to examine 46 cases of facial bone fractures, maxillary cysts, obstructive sleep apnea syndrome (OSAS) and maxillary bone tumours. Data were input into a navigator workstation to create images similar to those displayed by a fiberoptic endoscope. Various 3-D images were obtained when probe sites were changed. RESULTS: Virtual endoscopy can clearly display the anatomic structure of the paranasal sinuses, nasopharyngeal cavity and upper respiratory tract, revealing damage to the sinus wall caused by a bone tumor or fracture. In addition, the technique can be used to indicate volume changes of the upper respiratory tract in OSAS patients. These results were similar to those obtained through actual operations. CONCLUSIONS: Virtual endoscopy is a new method which produces very clear images. It is reliable to provide detailed information for optimal operative planning.

Facial reconstruction with combined facial, neck and occipital skin flap transplantation.

To treat large facial defect (more than 6 cm x 4 cm in diameter) or a wound with bone exposure to atmosphere by less-traumatic, easier-healing reconstruction method, a pedicle flap including facial, neck, posterior auricle and occipital skin flap was designed and transferred, one by one, to repair facial defect as well as other flap donor sites, but occipital skin flap was only used to cover posterior auricle area. After 2-3 years follow-up, well-healed skin flaps with good color, elasticity and sensation were observed in all 16 patients. It is concluded that this method is effective and practical.