Advancing brain-inspired computing with hybrid neural networks.

Brain-inspired computing, drawing inspiration from the fundamental structure and information-processing mechanisms of the human brain, has gained significant momentum in recent years. It has emerged as a research paradigm centered on brain-computer dual-driven and multi-network integration. One noteworthy instance of this paradigm is the hybrid neural network (HNN), which integrates computer-science-oriented artificial neural networks (ANNs) with neuroscience-oriented spiking neural networks (SNNs). HNNs exhibit distinct advantages in various intelligent tasks, including perception, cognition and learning. This paper presents a comprehensive review of HNNs with an emphasis on their origin, concepts, biological perspective, construction framework and supporting systems. Furthermore, insights and suggestions for potential research directions are provided aiming to propel the advancement of the HNN paradigm.

Adaptive Synaptic Scaling in Spiking Networks for Continual Learning and Enhanced Robustness.

Synaptic plasticity plays a critical role in the expression power of brain neural networks. Among diverse plasticity rules, synaptic scaling presents indispensable effects on homeostasis maintenance and synaptic strength regulation. In the current modeling of brain-inspired spiking neural networks (SNN), backpropagation through time is widely adopted because it can achieve high performance using a small number of time steps. Nevertheless, the synaptic scaling mechanism has not yet been well touched. In this work, we propose an experience-dependent adaptive synaptic scaling mechanism (AS-SNN) for spiking neural networks. The learning process has two stages: First, in the forward path, adaptive short-term potentiation or depression is triggered for each synapse according to afferent stimuli intensity accumulated by presynaptic historical neural activities. Second, in the backward path, long-term consolidation is executed through gradient signals regulated by the corresponding scaling factor. This mechanism shapes the pattern selectivity of synapses and the information transfer they mediate. We theoretically prove that the proposed adaptive synaptic scaling function follows a contraction map and finally converges to an expected fixed point, in accordance with state-of-the-art results in three tasks on perturbation resistance, continual learning, and graph learning. Specifically, for the perturbation resistance and continual learning tasks, our approach improves the accuracy on the N-MNIST benchmark over the baseline by 44% and 25%, respectively. An expected firing rate callback and sparse coding can be observed in graph learning. Extensive experiments on ablation study and cost evaluation evidence the effectiveness and efficiency of our nonparametric adaptive scaling method, which demonstrates the great potential of SNN in continual learning and robust learning.

Virulence, antimicrobial resistance, and molecular characteristics of carbapenem-resistant Klebsiella pneumoniae in a hospital in Shijiazhuang City from China.

Carbapenem-resistant Klebsiella pneumoniae (CRKP), as one of the most common drug-resistant bacteria threatening human health, is hyper-resistant to multiple antimicrobial drugs and carbapenems, which can be dealt with only limited clinical treatment options. This study described the epidemiological characteristics of CRKP in this tertiary care hospital from 2016 to 2020. Specimen sources included blood, sputum, alveolar lavage fluid, puncture fluid, secretions from a burn wound, and urine. Among the 87 carbapenem-resistant strains, ST11 was the predominant isolate, followed by ST15, ST273, ST340, and ST626. These STs were in broad agreement with the STs defined by pulsed-field gel electrophoresis clustering analysis in discriminating clusters of related strains. Most CRKP isolates contained the blaKPC-2 gene, some isolates carried the blaOXA-1, blaNDM-1, and blaNDM-5 genes, and the isolates carrying carbapenem resistance genes were more resistant to the antimicrobials of ß-lactams, carbapenems, macrolides, and fluoroquinolone. The OmpK35 and OmpK37 genes were detected in all CRKP strains, and the Ompk36 gene was detected in some CRKP strains. All detected OmpK37 had 4 mutant sites, and OmpK36 had 11 mutant sites, while no mutant sites were found in OmpK35. More than half of the CRKP strains contained the OqxA and OqxB efflux pump genes. The virulence genes were most commonly combined with urea-wabG-fimH-entB-ybtS-uge-ycf. Only one CRKP isolate was detected with the K54 podoconjugate serotype. This study elucidated the clinical epidemiological features and molecular typing of CRKP, and grasped the distribution of drug-resistant genotypes, podocyte serotypes, and virulence genes of CRKP, providing some guidance for the subsequent treatment of CRKP infection.

Enhancing spiking neural networks with hybrid top-down attention.

As the representatives of brain-inspired models at the neuronal level, spiking neural networks (SNNs) have shown great promise in processing spatiotemporal information with intrinsic temporal dynamics. SNNs are expected to further improve their robustness and computing efficiency by introducing top-down attention at the architectural level, which is crucial for the human brain to support advanced intelligence. However, this attempt encounters difficulties in optimizing the attention in SNNs largely due to the lack of annotations. Here, we develop a hybrid network model with a top-down attention mechanism (HTDA) by incorporating an artificial neural network (ANN) to generate attention maps based on the features extracted by a feedforward SNN. The attention map is then used to modulate the encoding layer of the SNN so that it focuses on the most informative sensory input. To facilitate direct learning of attention maps and avoid labor-intensive annotations, we propose a general principle and a corresponding weakly-supervised objective, which promotes the HTDA model to utilize an integral and small subset of the input to give accurate predictions. On this basis, the ANN and the SNN can be jointly optimized by surrogate gradient descent in an end-to-end manner. We comprehensively evaluated the HTDA model on object recognition tasks, which demonstrates strong robustness to adversarial noise, high computing efficiency, and good interpretability. On the widely-adopted CIFAR-10, CIFAR-100, and MNIST benchmarks, the HTDA model reduces firing rates by up to 50% and improves adversarial robustness by up to 10% with comparable or better accuracy compared with the state-of-the-art SNNs. The HTDA model is also verified on dynamic neuromorphic datasets and achieves consistent improvements. This study provides a new way to boost the performance of SNNs by employing a hybrid top-down attention mechanism.

Tsp2 Facilitates Tumor-associated Fibroblasts Formation and Promotes Tumor Progression in Retroperitoneal Liposarcoma.

Retroperitoneal liposarcoma (RLPS) is the most common subtype of retroperitoneal soft tissue sarcoma, characterized by a high recurrence rate and insensitivity to radiotherapy and chemotherapy. The function of tumor microenvironmental components, especially tumor-associated fibroblasts (TAFs), remains unclear in RLPS. The crosstalk between tumor cells and stromal cells should be clarified for therapy target discovery in RLPS. In this study, we demonstrated that TAFs from dedifferentiated liposarcoma (DDLPS) could attract LPS cells and promote their proliferation and migration. However, although α-SMA is positively expressed in RLPS, its expression does not indicate prognosis. By screening differentially expressed genes, performing Oncomine visualization, TCGA gene expression correlation analysis and qPCR verification, we determined that thrombospondin-2 (THBS2) gene expression was related to TAFs. The expression of Tsp2 protein, which was encoded by THBS2, was correlated with α-SMA expression, and it was an independent predictive factor for disease-free survival and recurrence-free survival in patients with RLPS. In vitro, Tsp2 facilitated the transformation of bone marrow-derived fibroblasts (BMFs) to TAFs and promoted the malignant biological behaviors of LPS cells by activating the MAPK/MEK/ERK pathway. Therefore, suppression of Tsp2 is expected to be a promising treatment method for RLPS patients.

Establishment and evaluation of retroperitoneal liposarcoma patient-derived xenograft models: an ideal model for preclinical study.

Retroperitoneal liposarcoma (RLPS) is one of the most common subtypes of retroperitoneal soft tissue sarcomas. It is characterized by poor sensitivity to radiotherapy and chemotherapy and a low success rate of complete surgical resection. However, there are few reliable preclinical RLPS models for target discovery and therapy research. In this study, we aimed to establish RLPS patient-derived xenograft (PDX) models that are useful for biological research and preclinical drug trials. A total of 56 freshly resected RLPS tissues were subcutaneously transplanted into non-obese diabetic-severe combined immune deficient (NOD-SCID) mice, with subsequent xenotransplantation into second-generation mice. The tumor engraftment rate of first generation PDXs was 44.64%, and higher success rates were obtained from implantations of dedifferentiated, myxous, pleomorphic, high-grade liposarcomas and those with retroperitoneal organ infiltration. The first- and second- generation PDX models preserved the histopathological morphology, gene mutation profiles and MDM2 amplification of the primary tissues. PDX models can also provide the benefit of retaining original tumor biology and microenvironment characteristics, such as abnormal adipose differentiation, elevated Ki67 levels, high microvessel density, cancer-associated fibroblast presence, and tumor-associated macrophage infiltration. Overall survival (OS) and disease-free survival (DFS) of patients with successful first-generation PDX engraftment were significantly poorer than those with failed engraftment. Treatment with MDM2 inhibitor RG7112 significantly suppressed tumor growth of DDLPS PDX in mice. In conclusion, we successfully established RLPS PDX models that were histologically, genetically, and molecularly consistent with the original tissues. These models might provide opportunities for advancing RLPS tumor biology research, facilitating the development of novel drugs, particularly those targeting MDM2 amplification, adipose differentiation process, angiogenesis, cancer-associated fibroblasts, and so on.

A framework for the general design and computation of hybrid neural networks.

There is a growing trend to design hybrid neural networks (HNNs) by combining spiking neural networks and artificial neural networks to leverage the strengths of both. Here, we propose a framework for general design and computation of HNNs by introducing hybrid units (HUs) as a linkage interface. The framework not only integrates key features of these computing paradigms but also decouples them to improve flexibility and efficiency. HUs are designable and learnable to promote transmission and modulation of hybrid information flows in HNNs. Through three cases, we demonstrate that the framework can facilitate hybrid model design. The hybrid sensing network implements multi-pathway sensing, achieving high tracking accuracy and energy efficiency. The hybrid modulation network implements hierarchical information abstraction, enabling meta-continual learning of multiple tasks. The hybrid reasoning network performs multimodal reasoning in an interpretable, robust and parallel manner. This study advances cross-paradigm modeling for a broad range of intelligent tasks.

Neuromorphic computing chip with spatiotemporal elasticity for multi-intelligent-tasking robots.

Recent advances in artificial intelligence have enhanced the abilities of mobile robots in dealing with complex and dynamic scenarios. However, to enable computationally intensive algorithms to be executed locally in multitask robots with low latency and high efficiency, innovations in computing hardware are required. Here, we report TianjicX, a neuromorphic computing hardware that can support true concurrent execution of multiple cross-computing-paradigm neural network (NN) models with various coordination manners for robotics. With spatiotemporal elasticity, TianjicX can support adaptive allocation of computing resources and scheduling of execution time for each task. Key to this approach is a high-level model, "Rivulet," which bridges the gap between robotic-level requirements and hardware implementations. It abstracts the execution of NN tasks through distribution of static data and streaming of dynamic data to form the basic activity context, adopts time and space slices to achieve elastic resource allocation for each activity, and performs configurable hybrid synchronous-asynchronous grouping. Thereby, Rivulet is capable of supporting independent and interactive execution. Building on Rivulet with hardware design for realizing spatiotemporal elasticity, a 28-nanometer TianjicX neuromorphic chip with event-driven, high parallelism, low latency, and low power was developed. Using a single TianjicX chip and a specially developed compiler stack, we built a multi-intelligent-tasking mobile robot, Tianjicat, to perform a cat-and-mouse game. Multiple tasks, including sound recognition and tracking, object recognition, obstacle avoidance, and decision-making, can be concurrently executed. Compared with NVIDIA Jetson TX2, latency is substantially reduced by 79.09 times, and dynamic power is reduced by 50.66%.

Anti-Tumor Effect of Apatinib and Relevant Mechanisms in Liposarcoma.

BACKGROUND: Primary retroperitoneal liposarcomas (RLPSs) are rare heterogeneous tumors for which there are few effective therapies. Certain anti-angiogenic tyrosine kinase inhibitors have demonstrated efficacy against various solid tumors. The aims of this study were to investigate the effect of Apatinib against retroperitoneal liposarcoma cells and its underlying mechanism and to explore the anti-tumor efficacy of a combination of Apatinib and Epirubicin. METHODS: CD34 immunohistochemical staining was used to measure microvessel density (MVD) in 89 retroperitoneal liposarcoma tissues. We used CCK-8 cell proliferation, clone formation, Transwell migration, invasion assays and flow cytometry to evaluate the effects of Apatinib alone and the combination of Apatinib and Epirubicin on liposarcoma cells. High-throughput RNA sequencing and western-blotting was used to identify key differentially expressed genes (DEGs) in SW872 cell line after application of Apatinib. Murine patient-derived tumor xenograft (PDX) was established to assess the efficacy and safety of Apatinib monotherapy and the combination of Apatinib and Epirubicin in RLPS. RESULTS: The microvessel density (MVD) varied widely among retroperitoneal liposarcoma tissues. Compared with the low-MVD group, the high-MVD group had poorer overall survival. Apatinib inhibited the liposarcoma cell proliferation, invasion and migration, increased the proportion of apoptosis, and induced G1 phase arrest. In addition, the combination of Apatinib and Epirubicin enhanced the foregoing inhibitory effects. High-throughput RNA sequencing showed that Apatinib downregulated the expression of TYMS and RRM2. Western blotting verified that Apatinib downregulated the TYMS/STAT3/PD-L1 pathway and inhibited liposarcoma proliferation by suppressing the RRM2/PI3K/AKT/mTOR pathway. In the murine PDX model of retroperitoneal liposarcoma, Apatinib and its combination with Epirubicin significantly inhibited microvessel formation and repressed tumor growth safely and effectively. CONCLUSIONS: Apatinib and its combination with Epirubicin showed strong efficacy against liposarcoma both in vitro and in vivo. Apatinib might inhibit liposarcoma cell proliferation through the RRM2/PI3K/AKT/mTOR signaling pathway and downregulate PD-L1 via the TYMS/STAT3 signaling pathway.

Adversarial symmetric GANs: Bridging adversarial samples and adversarial networks.

Generative adversarial networks have achieved remarkable performance on various tasks but suffer from training instability. Despite many training strategies proposed to improve training stability, this issue remains as a challenge. In this paper, we investigate the training instability from the perspective of adversarial samples and reveal that adversarial training on fake samples is implemented in vanilla GANs, but adversarial training on real samples has long been overlooked. Consequently, the discriminator is extremely vulnerable to adversarial perturbation and the gradient given by the discriminator contains non-informative adversarial noises, which hinders the generator from catching the pattern of real samples. Here, we develop adversarial symmetric GANs (AS-GANs) that incorporate adversarial training of the discriminator on real samples into vanilla GANs, making adversarial training symmetrical. The discriminator is therefore more robust and provides more informative gradient with less adversarial noise, thereby stabilizing training and accelerating convergence. The effectiveness of the AS-GANs is verified on image generation on CIFAR-10, CIFAR-100, CelebA, and LSUN with varied network architectures. Not only the training is more stabilized, but the FID scores of generated samples are consistently improved by a large margin compared to the baseline. Theoretical analysis is also conducted to explain why AS-GAN can improve training. The bridging of adversarial samples and adversarial networks provides a new approach to further develop adversarial networks.

Understanding chain folding morphology of semicrystalline polymers based on a rod-coil multiblock model.

We employ a rod-coil multiblock molecular chain model to investigate chain folding behavior, which is a significant characteristic in semicrystalline polymers, by using the method of self-consistent field theory (SCFT). Polymer chains with different conformations in crystalline and amorphous regions are described by rigid rod chains and flexible Gaussian chains, respectively. At present, we concentrate on the thermodynamic behaviors of polymer semi-crystals after the formation of the initial lamellar crystals. A new mechanism for lamellar thickening is proposed to realize that the end of lamellar thickening depends on the crystallinity degree. In other words, it is impossible for lamellae to develop into extended-chain crystals by means of lamellar thickening if crystallinity is limited to a certain degree. We further discuss the competition between crystalline and amorphous regions and its influence on crystallization behaviors, such as the formation of double lamellae, chain tilt, the anomalies and adjacent re-entry. The synergistic influences of the driving force of crystallization, interfacial energy and crystallinity degree on chain folding behavior are also investigated when the density anomalies in amorphous regions are excluded. Our model demonstrates advantages in accurately describing the mesoscopic layered structures of semicrystalline polymers based upon a microscopic chain model and provides at least a semi-quantitative thermodynamic picture for chain folding.

Effect of Geometrical Asymmetry on the Phase Behavior of Rod-Coil Diblock Copolymers.

The effect of geometrical asymmetry ß (described by the length-diameter ratio of rods) on the rod-coil diblock copolymer phase behavior is studied by implementation of self-consistent field theory (SCFT) in three-dimensional (3D) position space while considering the rod orientation on the spherical surface. The phase diagrams at different geometrical asymmetry show that the aspect ratio of rods ß influences not only the order-disorder transition (ODT) but also the order-order transition (OOT). By exploring the phase diagram with interactions between rods and coils plotted against ß, the ß effect on the phase diagram is similar to the copolymer composition f. This suggests that non-lamellae structures can be obtained by tuning ß, besides f. When the rods are slim compared with the isotropic shape of the coil segment (ß is relatively large), the phase behavior is quite different from that of coil-coil diblock copolymers. In this case, only hexagonal cylinders with the coil at the convex side of the interface and lamella phases are stable even in the absence of orientational interaction between rods. The phase diagram is no longer symmetrical about the symmetric copolymer composition and cylinder phases occupy the large area of the phase diagram. The ODT is much lower than that of the coil-coil diblock copolymer system and the triple point at which disordered, cylinder and lamella phases coexist in equilibrium is located at rod composition fR = 0.66. In contrast, when the rods are short and stumpy (ß is smaller), the stretching entropy cost of coils can be alleviated and the phase behavior is similar to coil-coil diblocks. Therefore, the hexagonal cylinder phase formed by coils is also found beside the former two structures. Moreover, the ODT may even become a little higher than that of the coil-coil diblock copolymers due to the large interfacial area per chain provided by the stumpy rods, thus compensating the stretching entropy loss of the coils.

Identification of recurrence-related serum microRNAs in hepatocellular carcinoma following hepatectomy.

Hepatocellular carcinoma (HCC) is one of the most deadly tumors. Prognosis of patients with HCC is generally poor due to the high recurrence rate. In the present study, TaqMan Real-time PCR microRNA Array was used to identify differentially expressed miRNAs from 10 tumor tissue samples (5 from recurrence group vs. 5 from non-recurrence group) and the matched serum samples. Four differentially expressed miRNAs (miR-486-5p, miR-422a, miR-125b and miR-139-5p) were further quantified in 20 tumor tissues and 116 HCC patients' serum before they received hepatectomy. Univariate analysis revealed that miR-486-5p, miR-422a and miR-125b were significantly associated with patients' relapse free survival (RFS). Multivariate analysis demonstrated that miR-486-5p, AFP and microvascular invasion (MVI) were the independent prognostic factors associated with RFS in this cohort (p = 0.000, 0.043, 0.000, respectively). Besides, the expression levels of miR-486-5p were positively correlated in tumor tissues and the paired serum samples, so was miR-422a. The probability of the prognostic accuracy of miR-486-5p in predicting postoperative recurrence of HCC within the first year was 76.79% (65.38% specificity and 81.58% sensitivity), which was almost equal to the classifier established by combination of AFP and MVI (75.98% probability, 63.13% specificity and 85.90% sensitivity). Furthermore, the combination of AFP, MVI and miR-486-5p yielded a ROC curve area of 88.02% (69.20% specificity and 92.10% sensitivity). Our study was the first to identify that serum miR-486-5p could be used to stratify the patients with higher recurrence risk before hepatic resection and potentially guide more effective surveillance strategies for them.

Prognostic Significance of Preoperative Serum Alpha- fetoprotein in Hepatocellular Carcinoma and Correlation with Clinicopathological Factors: a Single-center Experience from China.

OBJECTIVES: To investigate the prognosis significance of preoperative serum alpha-fetoprotein (AFP) and the correlation with clinicopathological factors of hepatocellular carcinoma (HCC) patients who underwent hepatectomy. MATERIALS AND METHODS: Clinicopathological data of retrospective analysis were collected for 251 HCC patients undergoing hepatectomy in this study. According to preoperative AFP level, patients were categorized into AFP-negative (0-20 ng/mL) and AFP-positive (>20 ng/mL) groups for Kaplan-Meier analysis and Cox proportional hazard regression modeling. RESULTS: The results demonstrated that increased AFP was associated with longer prothrombin time (PTs), liver capsule invasion, low grade differentiation, and late Barcelona Clinic Liver Center (BCLC) stage. Moreover, the female patients had a greater prevalence of increased preoperative AFP than male patients [284.8 (3.975-3167.5) vs (3.653-140.65); Z-2.895, p=0.004]. The 1-, 3-, and 5-year recurrence-free survival (RFS) rates were 78.1, 57.5, and 40.6 % in the AFP-negative group and 61.8, 37.7, and 31.4 %, respectively, in the AFP-positive group (log-rank test 8.312, p=0.004). The 1-, 3-, and 5-year overall survival (OS) rates were 94.4, 83.8, and 62.3% in the AFP-negative group and 87.2, 60.0, and 36.7%, respectively, in the AFP-positive group. The difference was statistically significant (log-rank test, 16.884, p=0.000). Cox proportional-hazards model identified preoperative AFP to be an independent prognostic predictor of overall survival. CONCLUSIONS: Preoperative serum AFP is an independent predictor of prognosis among HCC patients following surgical resection. Female patients have a higher preoperative AFP than their male counterparts.

Hepatocellular adenoma with malignant transformation in male patients with non-cirrhotic livers.

INTRODUCTION: Hepatocellular adenomas (HCAs), with a risk of malignant transformation into hepatocellular carcinoma (HCC), classically develop in young women who are taking oral contraceptives. It is now clear that HCAs may also occur in men. However, it is rarely reported that HCAs with malignant transformation occur in male patients with non-cirrhotic livers. This study aimed to characterize the malignancy of HCAs occurring in male patients. METHODS: All patients with HCAs with malignant transformation who underwent hepatectomy at the Cancer Institute and Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College between January 1, 1999 and December 31, 2011 were enrolled in the study. The clinical characteristics as well as radiologic and pathologic data were reviewed. RESULTS: HCAs with malignant transformation were observed in 5 male patients with non-cirrhotic livers, but not in female patients. The alpha-fetoprotein (AFP) levels were higher in patients with HCAs with malignant transformation than in patients with HCAs without malignant transformation. The diameters of the tumors with malignant transformation were larger than 5 cm in 3 cases and smaller than 5 cm in 2 cases. The 5 patients were all alive without recurrence by the end of the study period. The disease-free survival times of the 5 patients were 26, 48, 69, 69, and 92 months. CONCLUSION: Our results indicate that resection would be advised even if the presumptive diagnosis is adenoma smaller than 5 cm in diameter, especially in male patients.

Primary lymphoepithelioma-like hepatocellular carcinoma: report of a locally advanced case and review of literature.

Lymphoepithelioma-like hepatocellular carcinoma is rare, which has been recognized as a variant of hepatocellular carcinoma. Here we report a locally advanced case of lymphoepithelioma-like hepatocellular carcinoma. A 50-year-old man with chronic hepatitis B virus infection presented with a single mass in the liver and two enlarged lymph nodes in retroperitoneum suspected to be hepatocellular carcinoma with lymph node metastasis. After discussion by multidisciplinary team, the patient underwent hepatectomy of VIII segment and dissection of two enlarged lymph nodes. One month after the operation, pre-chemotherapy abdominal computed tomography (CT) showed retroperitoneal enlarged lymph nodes, considered as local recurrence. Therefore, 3 cycles of oxaliplatin and tegafur gimeracil oteracil potassium capsule and 3 cycles of paclitaxel and cisplatin were offered, and post-chemotherapy abdominal CT revealed disease remained stable. The patient has been alive for 6 months since performance of surgery. Our report suggests that even locally advanced lymphoepithelioma-like hepatocellular carcinoma may have a good prognosis and operation and postoperative chemotherapy may benefit the patient.

[Initial comparison of regional ischemic preconditioning and hemi-hepatic vascular inflow occlusion in resection of hepatocellular carcinoma].

OBJECTIVE: To evaluate preliminarily the clinical efficacy of two types of hepatic inflow occlusion in hepatectomy for hepatocellular carcinoma (HCC). METHODS: A total of 54 patients with HCC who underwent hepatectomy were divided into two groups: RIP group (regional ischemic preconditioning with continuous clamping, n=15) and HHV group (hemi-hepatic vascular inflow occlusion, n=39). HHV was performed by placing a clamp on the right hepatic artery and right portal vein, and was maintained until the liver resection was completed. In the RIP group, HHV was preceded by a 5-min period of ischemia followed by 5 min of reperfusion. The clinical indicators of the two groups were compared. RESULTS: The volume of intraoperative blood loss had significant difference between the two groups (P=0.039). One case (6.7%) in the RIP group and 17 cases (43.6%) in the HHV group received postoperative blood transfusion, showing a significant difference (P=0.010). No postoperative 30-day mortality happened in all patients. No significant differences were found between the two groups in hospital stay or postoperative morbidity, including hepatic insufficiency, infection, ascites, pleural effusion, cardiopulmonary complications and intestinal ventilation time (P>0.05 for all).The RIP group had a significantly higher PTA level at postoperative days 3 and 5 (P<0.001). Although no significant differences were found between the two groups regarding total bilirubin, albumin, prealbumin and aminotransferase (P>0.05) during any postoperative stage, the ALT recovered to normal level in 5 patients (33.3%) of the RIP group and only in one case (2.7%) of the HHV group, with a significant difference between the two groups (P=0.006). CONCLUSION: The results of this study indicate that regional ischemic preconditioning may have better hemostatic effect on hepatectomy, can reduce postoperative blood transfusion and promote early recovery of liver function than hemi-hepatic vascular inflow occlusion.

Three-dimensional morphometric analysis for hepatectomy of centrally located hepatocellular carcinoma: a pilot study.

AIM: To describe a three-dimensional model (3DM) to accurately reconstruct anatomic relationships of centrally located hepatocellular carcinomas (HCCs). METHODS: From March 2013 to July 2014, reconstructions and visual simulations of centrally located HCCs were performed in 39 patients using a 3D subject-based computed tomography (CT) model with custom-developed software. CT images were used for the 3D reconstruction of Couinaud's pedicles and hepatic veins, and the calculation of corresponding tumor territories and hepatic segments was performed using Yorktal DMIT software. The respective volume, surgical margin, and simulated virtual resection of tumors were also estimated by this model preoperatively. All patients were treated surgically and the results were retrospectively assessed. Clinical characteristics, imaging data, procedure variables, pathologic features, and postoperative data were recorded and compared to determine the reliability of the model. RESULTS: 3D reconstruction allowed stereoscopic identification of the spatial relationships between physiologic and pathologic structures, and offered quantifiable liver resection proposals based on individualized liver anatomy. The predicted values were consistent with the actual values for tumor mass volume (82.4 ± 109.1 mL vs 84.1 ± 108.9 mL, P = 0.910), surgical margin (10.1 ± 6.2 mm vs 9.1 ± 5.9 mm, P = 0.488), and maximum tumor diameter (4.61 ± 2.16 cm vs 4.53 ± 2.14 cm, P = 0.871). In addition, the number and extent of portal venous ramifications, as well as their relation to hepatic veins, were visualized. Preoperative planning based on simulated resection facilitated complete resection of large tumors located in the confluence of major vessels. And most of the predicted data were correlated with intraoperative findings. CONCLUSION: This 3DM provides quantitative morphometry of tumor masses and a stereo-relationship with adjacent structures, thus providing a promising technique for the management of centrally located HCCs.

[Effect of resection margin and tumor number on survival of patients with small liver cancer].

OBJECTIVE: To explore the significance of resection margin and tumor number on survival of patients with small liver cancer after hepatectomy. METHODS: We collected 219 cases with small liver cancer undergoing hepatectomy in Cancer Hospital, Chinese Academy of Medical Sciences between December 2003 to July 2013. The survival rates were compared by log-rank test between two resection margin groups (≥ 1 cm vs. <1 cm), different tumor number groups (single tumor vs. multiple tumors). We also performed a multifactor analysis by Cox model. RESULTS: The 1-, 3-, 5- and 10- year overall survival rates were 95.9%, 85.3%, 67.8% and 53.3%, respectively, in all patients. The median survival time was 28 months in the group of <1 cm resection margin and 36 months in the group of ≥ 1 cm resection margin (P=0.249). The median survival time was 36 months in the group of single tumor and 26 months in the group of multiple tumors (P=0.448). The multifactor analysis also did not show significant effect of resection margin and tumor number on the patients' survival. CONCLUSIONS: For small liver cancer, the resection margin of 1 cm might be advised. Increasing resection margin in further could probably not improve therapeutic effect. Standardized operation and combined treatment will decrease the negative influence of multiple tumors on overall survival.

[Effect of preoperative transcatheter arterial chemoembolization on the perioperative outcome of patients with hepatocellular carcinoma].

OBJECTIVE: To explore the surgical risk, perioperative outcome and the response of patients with hepatocellular carcinoma (HCC) after preoperative transcatheter arterial chemoembolization (TACE). METHODS: A retrospective case-matched study was conducted to compare the characteristics and corresponding measures of patients in the preoperative TACE group and the control group without TACE. A total of 105 patients (82 patients with selective and dynamic region-specific vascular occlusion to perform hepatectomy for patients with complex hepatocellular carcinoma) was included in this study, in which 35 patients underwent TACE therapy, and a 1:2 matched control group of 70 subjects. RESULTS: The patients of preoperative TACE therapy group had a higher level of γ-glutamyl transpeptidase before operation (119.52±98.83) U/L vs. (67.39±61.25) U/L (P=0.040). The operation time was longer in the TACE group than that in the control group but with a non-significant difference (232.60±95.43) min vs. (218.70±75.13) min (P=0.052). The postoperative recovery of liver function and severe complications in the preoperative TACE group were similar to that in the control group (P>0.05). There were no massive hemorrhage, biliary fistula and 30-d death neither in the treatment group and matched control group. CONCLUSIONS: Preoperative TACE therapy has certain negative effect on liver function. It is preferable to use selective and dynamic region-specific vascular occlusion technique during hepatectomy and combine with reasonable perioperative treatment for this group of patients, that can ensure safety of patients and promote their rapid recovery.