Background and Aim: Treatment strategy for hepatocellular carcinoma (HCC) and Vp4 [main trunk] portal vein tumor thrombosis (PVTT) remains limited due to posttreatment liver failure. We aimed to assess the efficacy of irradiation stent placement with ¹²⁵I plus transcatheter arterial chemoembolization (TACE) (ISP-TACE) compared to sorafenib plus TACE (Sora-TACE) in these patients. Methods: In this multicenter randomized controlled trial, participants with HCC and Vp4 PVTT without extrahepatic metastases were enrolled from November 2018 to July 2021 at 16 medical centers. The primary endpoint was overall survival (OS). The secondary endpoints were hepatic function, time to symptomatic progression, patency of portal vein, disease control rate, and treatment safety. Results: Of 105 randomized participants, 51 were assigned to the ISP-TACE group, and 54 were assigned to the Sora-TACE group. The median OS was 9.9 months versus 6.3 months (95% CI: 0.27–0.82; P=0.01). Incidence of acute hepatic decompensation was 16% (8 of 51) versus 33% (18 of 54) (P=0.036). The time to symptomatic progression was 6.6 months versus 4.2 months (95% CI: 0.38–0.93; P=0.037). The median stent patency was 7.2 months (interquartile range, 4.7–9.3) in the ISP-TACE group. The disease control rate was 86% (44 of 51) versus 67% (36 of 54) (P=0.018). Incidences of adverse events at least grade 3 were comparable between the safety populations of the two groups: 16 of 49 (33%) versus 18 of 50 (36%) (P=0.73). Conclusion: Irradiation stent placement plus TACE showed superior results compared with sorafenib plus TACE in prolonging OS in patients with HCC and Vp4 PVTT.

Purpose: Transarterial chemoembolization (TACE) was commonly applied in hepatocellular carcinoma (HCC) patients across BCLC A-C stages with heterogeneous outcomes in real-world practice. We aimed to develop a neutrophil-to-lymphocyte ratio (NLR) and sarcopenia-based prognostic nomogram to estimate the prognosis of HCC patients after TACE treatment. Patients and Methods: Between June 2013 and December 2019, a total of 364 HCC patients who underwent TACE were included and randomly assigned to the training (n=255) and the validation cohort (n=109). Sarcopenia was diagnosed based on the third lumbar vertebra skeletal muscle mass index (L3-SMI). The multivariate Cox proportional risk model was used to generate a nomogram. Results: NLR ≥ 4.0, sarcopenia, alpha-fetoprotein (AFP) ≥ 200 ng/mL, albumin-bilirubin (ALBI) grade 2 or 3, number of lesions (≥ 2), and maximum size of the lesion (≥ 5 cm) were independent predictors for overall survival (OS) (P < 0.05). The calibration curve shows that the predicted results agree well with the observed results. The time-dependent areas under the receiver-operating characteristic curves for OS at 1, 2, and 3 years predicted by the nomogram were 0.818/0.827, 0.742/0.823, and 0.748/0.836 in both training and validation cohorts. Nomogram can divide patients into low-, medium- and high-risk groups based on predictor factors. The C-indexes of the nomogram for OS were 0.782/0.728 in the training and validation cohorts, outperforming other currently available models. Conclusion: A novel nomogram based on NLR and sarcopenia may be useful to predict the prognosis of HCC patients who underwent TACE across BCLC A-C stage patients.

The development of a new generation of minimally invasive surgery is mainly reflected in robot-assisted diagnosis and treatment methods and their clinical applications. It is a clinical concern for robot-assisted surgery to use a multi-joint robotic arm performing human ultrasound scanning or ultrasound-guided percutaneous puncture. Among them, the motion control of the robotic arm, and the guiding and contact scanning processes of the ultrasonic (US-) probe determine the diagnosis effect, as well as the accuracy and safety of puncture surgery. To address these challenges, this study developed an intelligent robot-assisted system integrating autonomous US inspection and needle positioning, which has relation to several intelligent algorithms such as adaptive flexible control of the robot arm, autonomous US-scanning, and real-time attitude adjustment of the puncture needle. To improve the cooperativity of the spatial operation of the robot end-effector, we propose an adaptive flexible control algorithm that allows the operator to control the robot arm flexibly with low damping. To achieve the stability and uniformity of contact detection and imaging, we introduced a self-scanning method of US-probe based on reinforcement learning and built a software model of variable stiffness based on MuJoco to verify the constant force and velocity required by the end mechanism. We conducted a fixed trajectory scanning experiment at a scanning speed of 0.06 m/s. The force curve generally converges towards the desired contact force of 10 N, with minor oscillations around this value. For surgical process monitoring, we adopted the puncture needle detection algorithm based on Unet++ to acquire the position and attitude information of the puncture needle in real time. In short, we proposed and verified an adaptive control method and learning strategy by using an UR robotic arm equipped with a US-probe and puncture needle, and we improved the intelligence of the US-guided puncture robot.