2020年 · 第100卷第39期

normal pressure hydrocephalus (NPH) refers to a group of clinical syndromes with gait instability, cognitive dysfunction and urinary incontinence as typical clinical manifestations, ventricular enlargement on imaging examination, and cerebrospinal fluid pressure measurement within the normal range, which can be divided into idiopathic and secondary. Cerebrovascular disease-related normal NPH is a special type of NPH secondary to various cerebrovascular diseases^{[1,2,3,4,5,6]}It is not uncommon in clinical practice, often masked by the primary disease, and clinicians fail to detect it in time, resulting in poor prognosis for patients. So far, the clinical experience on cerebrovascular disease-related NPH is relatively lacking, and the diagnosis and treatment are also lacking standards. The purpose of this consensus is to provide a reference for neurosurgeons at home and abroad in the diagnosis and treatment of cerebrovascular disease-related NPH.

With the increase of the elderly population in China and the continuous improvement of the standardized level of thrombotic disease management, the proportion of patients taking antithrombotic drugs for a long time before surgery in surgical patients is also increasing. Antithrombotic drugs mainly include anticoagulants, antiplatelet drugs and fibrinolytic drugs^[1]。 For patients who take antithrombotic drugs (anticoagulants and antiplatelet drugs) for a long time before surgery, continuing to take them during the perioperative period can increase the risk of surgical bleeding, while discontinuing them can increase the risk of thromboembolic events. Given that hemorrhage and thromboembolism have important adverse effects on patients' near-and long-term prognosis, how to balance the two risks is crucial, and sometimes multidisciplinary collaborative evaluation is required to develop individualized patient management protocols. Combined with the latest evidence and guidelines at home and abroad, and through the discussion of relevant multidisciplinary experts, this consensus forms recommendations for the thromboembolic risk classification of common thromboembolic diseases in internal medicine, the risk classification of surgical bleeding, the withdrawal timing and bridging plan of antithrombotic drugs, the management of antithrombotic drugs during the perioperative period of elective and emergency surgery, and the management of regional block anesthesia and antithrombotic drugs, so as to reduce the incidence of major cardiovascular and cerebrovascular events and adverse hemorrhagic events during the perioperative period of such patients and ensure the safety of patients during the perioperative period.

Noninvasive assessment of hepatic fibrosis is a hot topic in clinical research at present. Transient elastography, shear wave elastography, and magnetic resonance (MRI) elastography are increasingly used. Deep convolutional neural networks (DCNN) for CT and MRI images to evaluate liver fibrosis showed high diagnostic efficacy, while deep learning for B-mode ultrasound images was less reported. This study predicted liver fibrosis METAVIR score by establishing a deep convolutional neural network on B-mode ultrasound images. 13 608 ultrasound images from 3 446 patients in referral centers of two specialized institutions were collected for training DCNN. Patients underwent surgical resection, liver puncture biopsy or transient elastography, and the estimated METAVIR scores derived from pathological specimens or transient elastography were used as reference standards to establish a four-category model (F0 vs F1 vs F23 vs F4). The internal test set included 266 patients with 300 images and the external test set included 572 patients with 1,232 images. The results showed that the diagnostic accuracy of the four-category model in the internal and external test sets was 83.5% and 76.4%, respectively. The area under the operating characteristic curve (AUC) of subjects diagnosed with cirrhosis (F4) was 0.901 (95%) for the internal test set, respectivelyCI: 0.865~0.937), external test set 0.857 (95%CI：0.825~0.889）。 The AUC (0.857) of the DCNN used for the diagnosis of liver cirrhosis from the external test set was significantly higher than the diagnostic AUC of the 5 sonographers (AUC range 0.656 to 0.816;P<0.05）。

Stage T2 gallbladder cancer is further subdivided into stage T2a (peritoneal side) or stage T2b (hepatic side) depending on tumor location. Previous studies have suggested that liver resection can be avoided without affecting prognosis for peritoneal (T2a) tumors, but there is still a lack of reliable evidence. In order to explore the prognostic value of tumor location in stage T2 gallbladder cancer and to clarify the appropriate range of surgical resection, the study analyzed the clinical and pathological data of 937 patients undergoing surgery for gallbladder cancer in 14 hospitals in Korea, Japan, Chile and the United States. The results showed that the overall 5-year disease-free survival rate of patients was 70.6%, T2a was 74.5%, and T2b was 65.5% (P =0.028）。 Compared with cholecystectomy alone, extended cholecystectomy (gallbladder combined with liver resection) had a better 5-year disease-free survival rate (73.0% vs 61.5%;P =0.012）。 At T2a (76.5% vs. 66.1%;P =0.094) and T2b (68.2% vs. 56.2%;P =0.084) stage gallbladder cancer, the 5-year disease-free survival rate of extended cholecystectomy was slightly better than that of cholecystectomy alone. For patients undergoing extended cholecystectomy, 5-year disease-free survival was similar, including bedside liver wedge resection of the gallbladder and segment IVb/V liver resection (74.1% vs. 71.5%;P =0.720）。 Multivariate analysis indicated that the presence of clinical symptoms (HR=1.52，P =0.002), R1 resection (HR=1.96，P =0.004) and N1/N2 state (N1:HR=3.40，P<0.001； N2:HR=9.56，P<0.001) was an independent risk factor for recurrence. Among the recurrent patients, 70.8% had distant metastases. For stage T2 gallbladder cancer, tumor location is not an independent prognostic factor, and extended cholecystectomy is slightly better than simple cholecystectomy, so radical surgery should include liver resection and adequate lymph node dissection.