Background/Aims: Despite the high efficacy of direct-acting antivirals (DAAs), approximately 1–3% of hepatitis C virus (HCV) patients fail to achieve a sustained virological response. We conducted a nationwide study to investigate risk factors associated with DAA treatment failure. Machine-learning algorithms have been applied to discriminate subjects who may fail to respond to DAA therapy. Methods: We analyzed the Taiwan HCV Registry Program database to explore predictors of DAA failure in HCV patients. Fifty-five host and virological features were assessed using multivariate logistic regression, decision tree, random forest, eXtreme Gradient Boosting (XGBoost), and artificial neural network. The primary outcome was undetectable HCV RNA at 12 weeks after the end of treatment. Results: The training (n=23,955) and validation (n=10,346) datasets had similar baseline demographics, with an overall DAA failure rate of 1.6% (n=538). Multivariate logistic regression analysis revealed that liver cirrhosis, hepatocellular carcinoma, poor DAA adherence, and higher hemoglobin A1c were significantly associated with virological failure. XGBoost outperformed the other algorithms and logistic regression models, with an area under the receiver operating characteristic curve of 1.000 in the training dataset and 0.803 in the validation dataset. The top five predictors of treatment failure were HCV RNA, body mass index, α-fetoprotein, platelets, and FIB-4 index. The accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of the XGBoost model (cutoff value=0.5) were 99.5%, 69.7%, 99.9%, 97.4%, and 99.5%, respectively, for the entire dataset. Conclusions Machine learning algorithms effectively provide risk stratification for DAA failure and additional information on the factors associated with DAA failure.

Zuojinwan originated from Danxi′s Experiential Therapy （《丹溪心法》） in the Yuan dynasty. It is a representative prescription for the treatment of liver fire invading stomach syndrome， and is also one of the typical prescriptions of the anti-adjuvant method of traditional Chinese medicine （TCM）. In this paper， the method of bibliometrics was used to systematically sort out the ancient books of Zuojinwan， and 729 relevant literature data were obtained. After certain retrieval and screening， 57 ancient books of TCM were finally obtained. The statistics and analysis were carried out from the aspects of prescription source， historical evolution， composition， functions， evolution of prescription meaning， prescription dose， and preparation and usage of Zuojinwan. It was found that Zuojinwan was composed of Coptis chinensis rhizoma and Euodia rutaecarpa fructus in a ratio of 6∶1. It was mainly used for the treatment of liver fire invading stomach syndrome. The symptoms included pain in chest and hypochondrium， vomiting and bitter mouth， noisy acid-swallowing， red tongue coating yellow， and pulse string number. Later medical records recorded that Zuojinwan was mostly consistent with the original prescription. It mainly treated various diseases caused by liver fire， including left by liver fire， including left hypochondriac pain， swallowing acid and vomiting acid， tendon hernia and lump， epigastric pain， bitter mouth pulse string， head pain， diarrhea， gonorrhea， cold and hot， abdominal pain， alcohol wet yellowing， silence of oral dysentery and so on. There was little controversy in the analysis of relevant prescriptions. In the past dynasties， pills was mainly used， which was consistent with the original prescription. In modern times， it is mainly water flooding for pills or steamed cakes for pills， warm boiling water to serve 6 g， taking 2-3 g per time， the history is basically the same. In this paper， through the excavation， collation and systematic analysis of the ancient literature of TCM that recorded Zuojinwan， we hope to provide the literature basis for the development， inheritance and utilization of this famous classical formulas.

Objective:To establish and apply a new practical analytical method for identifying the fishy odor of Cordyceps based on headspace-solid phase microextraction-gas chromatography-triple quadrupole mass spectrometry （HS-SPME/GC-QQQ-MS/MS） technique. Method:The InertCap Pure-WAX capillary column （0.25 mm×30 m， 0.25 μm） was used for chromatographic separation. The injection port temperature was set at 250 ℃. The injection mode was split injection with a ratio of 5∶1. High purity helium was used as the carrier gas and control mode was set to constant pressure. The column flow rate was 1.43 mL∙min^-1， the linear velocity was 43.3 cm∙s^-1， and the purge flow rate was 3.0 mL∙min^-1. The chromatographic column temperature program as follows：maintained the initial temperature at 50 ℃ for 5 min， and increased the temperature at a rate of 10 ℃∙min^-1 to 250 ℃， held for 10 min. The column equilibrium time was 2.0 min. The ion source of mass spectrographic analysis was electron ionization with ion source temperature of 200 ℃， and the monitoring mode was set to multiple reaction monitoring. Result:Seven batches of Cordyceps samples were collected， including 3 batches from Sichuan， 3 batches from Qinghai and 1 batch from Tibet. There were six batches of counterfeits， including 3 batches from Sichuan， 2 batches from Guizhou and 1 batch in Xinjiang. A total of 81 volatile compounds were screened out in Cordyceps， which could be divided into 13 types （esters， ketones， aldehydes and others） according to the compound structure， indicating that the fishy odor of Cordyceps was a complex odor. There was no significant difference in the types of volatile compounds of Cordyceps from different regions， which suggested that these volatile compounds in Cordyceps produced in Tibet （Naqu）， Qinghai （Yushu and Guoluo） and Sichuan （Litang， Rangtang and Seda） were relatively consistent. However， the contents of some volatile compounds in Cordyceps produced in different regions were quite different， and 16 volatile compounds with significant difference were screened out， including 1-methoxy-2-propyl acetate， γ-octalactone， hexyl acetate and others， those compounds maybe could been used as the quality markers for identification of regions of Cordyceps. There was a large difference in volatile compounds between Cordyceps and its counterfeits， and 34 volatile compounds were screened out， including ethyl acetate， acetophenone， 2-ethyl-1-hexanol and others， those compounds maybe could been used as the quality markers for authenticity identification of Cordyceps. Conclusion:In summary， the established method for identifying the fishy odor of Cordyceps in this paper has the characteristics of high sensitivity， accuracy and simplicity， which can provide reference for the analysis of volatile compounds in other Chinese herbal medicines.

Background and objectiveIt is so far not clear that how myasthenia gravis (MG) affected the prognosis of thymoma patients. The aim of this assay is to compare the postoperative survival between patients with thymoma only and those with both thymoma and MG.MethodsThe Chinese Alliance for Research in Thymomas (ChART) registry recruited patients with thymoma from 18 centers over the country on an intention to treat basis from 1992 to 2012. Two groups were formed according to whether the patient complicated MG. Demographic and clinical data were reviewed, Patients were fol-lowed and their survival status were analyzed.Results There were 1,850 patients included in this study, including 421 with and 1,429 without MG. Complete thymectomy were done in 91.2% patients in MG group and 71.0% in non-MG group (P<0.05). There were more percentage of patients with the histology of thymoma AB, B1, or B2 (P<0.05) in MG group, and more percentage of patients with MG were in Masaoka stage I and II. The 5 year and 10 year OS rates were both higher in MG group (93%vs 88%; 83%vs 81%,P=0.034) respectively. The survival rate was signiifcantly higher in patients with MG when the Masaoka staging was III/IV (P=0.003). Among patients with advanced stage thymoma (stage III, IVa, IVb), the constitu-ent ratios of III, IVa, IVb were similar between MG and Non-MG group. Histologically, however, there were signiifcantly more proportion of AB/B1/B2/B3 in the MG group while there were more C in the non-MG group (P=0.000). Univariate analyses for all patients showed that MG, WHO classiifcation, Masaoka stage, surgical approach, chemotherapy and radiotherapy and resectability were signiifcant factors, and multivariate analysis showed WHO Classiifcation, Masaoka stage, and resectability were strong independent prognostic indicators.ConclusionAlthough MG is not an independent prognostic factor, the sur-vival of patients with thymoma was superior when MG was present, especially in late Masaoka stage patients. Possible reasons included early diagnosis of the tumor, better histologic types, an overall higher R0 resection and less recurrence.

Background and objectiveTo compare the predictive effect of the Masaoka-Koga staging system and the International Association for the Study of Lung Cancer (IASLC)/the International Thymic Malignancies Interest Group (ITMIG) proposal for the new TNM staging on prognosis of thymic malignancies using the Chinese Alliance for Research in Thymomas (ChART) retrospective database.MethodsFrom 1992 to 2012, 2,370 patients in ChART database were ret-rospectively reviewed. Of these, 1,198 patients with complete information on TNM stage, Masaoka-Koga stage, and survival were used for analysis. Cumulative incidence of recurrence (CIR) was assessed in R0 patients. Overall survival (OS) was evalu-ated both in an R0 resected cohort, as well as in all patients (any R status). CIR and OS were ifrst analyzed according to the Masaoka-Koga staging system. Then, they were compared using the new TNM staging proposal.Results Based on Masaoka-Koga staging system, signiifcant difference was detected in CIR among all stages. However, No survival difference was revealed between stage I and II, or between stage II and III. Stage IV carried the highest risk of recurrence and worst survival. According to the new TNM staging proposal, CIR in T1a was signiifcantly lower comparing to all other T categories (P<0.05) and there is a signiifcant difference in OS between T1a and T1b (P=0.004). T4 had the worst OS comparing to all other T categories. CIR and OS were signiifcantly worse in N(+) than in N0 patients. Signiifcant difference in CIR and OS was detected between M0 and M1b, but not between M0 and M1a. OS was almost always statistically different when comparison was made between stages I-IIIa and stages IIIb-IVb. However, no statistical difference could be detected among stages IIIb to IVb.Conclusion Compared with Masaoka-Koga staging, the IASLC/ITMIG TNM staging proposal not only describes the extent of tumor invasion but also provides information on lymphatic involvement and tumor dissemination. Further study using prospectively recorded information on the proposed TNM categories would be helpful to better grouping thymic tumors for predicting prognosis and guiding clinical management.

Background and objectiveTo evaluate the role of preoperative induction therapy on prognosis of local-ly advanced thymic malignancies.MethodsBetween 1994 and 2012, patients received preoperative induction therapies (IT group) in the Chinese Alliance for Research in Thymomas (ChART) database, were compared with those having surgery di-rectly atfer preoperative evaluation (DS group). All tumors receiving induction therapies were locally advanced (clinically stage III-IV) before treatment and those turned out to be in pathological stage I and II were considered downstaged by induction. Clinical pathological characteristics were retrospectively analyzed. To more accurately study the effect of induction therapies, stage IV patients were then excluded. Only stage I-III tumors in the IT group and stage III cases in the DS group were selected for further comparison in a subgroup analysis.Results Only 68 (4%) out of 1,713 patients had induction therapies, with a R0 resection of 67.6%, 5-year recurrence of 44.9%, and 5- and 10-year overall survivals (OS) of 49.7% and 19.9%. Seventeen pa-tients (25%) were downstaged atfer induction. Signiifcantly more thymomas were downstaged than thymic carcinomas (38.7%vs 13.9%,P=0.02). Tumors downstaged atfer induction had signiifcantly higher 5-year OS than those not downstaged (93.8%vs 35.6%,P=0.013). For the subgroup analysis when stage IV patients were excluded, 5-year OS was 85.2% in the DS group and 68.1% in the IT group (P<0.001), although R0 resection were similar (76.4%vs 73.3%,P=0.63). However, 5-year OS in tumors downstaged atfer induction (93.8%) was similar to those in the DS group (85.2%,P=0.438), both signiifcantly higher than those not downstaged atfer induction (35.6%,P<0.001).ConclusionOnly 68 (4%) out of 1,713 patients had induction therapies, with a R0 resection of 67.6%, 5-year recurrence of 44.9%, and 5- and 10-year overall survivals (OS) of 49.7% and 19.9%. Seventeen patients (25%) were downstaged atfer induction. Signiifcantly more thymomas were downstaged than thy-mic carcinomas (38.7%vs 13.9%,P=0.02). Tumors downstaged atfer induction had signiifcantly higher 5-year OS than those not downstaged (93.8%vs 35.6%,P=0.013). For the subgroup analysis when stage IV patients were excluded, 5-year OS was 85.2% in the DS group and 68.1% in the IT group (P<0.001), although R0 resection were similar (76.4%vs 73.3%,P=0.63). However, 5-year OS in tumors downstaged atfer induction (93.8%) was similar to those in the DS group (85.2%,P=0.438), both signiifcantly higher than those not downstaged atfer induction (35.6%,P<0.001).

Background and objectiveVideo-assisted thoracoscopic surgery (VATS) theoretically offers advantages over open thymectomy for clinically early-stage (Masaoka-Koga stage I and II) thymic malignancies. However, longterm outcomes have not been well studied. We compared the postoperative outcomes and survival from a cohort study based on the database of the Chinese Alliance for Research in Thymomas (ChART).MethodsBetween 1994 and 2012, data of 1,117 patients hav-ing surgery for clinically early-stage (Masaoka-Koga stage I and II) tumors were enrolled for the study. Among them, 241 cases underwent VATS thymectomy (VATS group), while 876 cases underwent open thymectomy (Open group). Univariate analyses were used to compare the clinical character and perioperative outcomes between the two groups. And multivariate analysis was performed to determine the independent predictive factors for long-term survival.Results Compared with the Open group, the VATS group had higher percentage of total thymectomy (80.5%vs 73.9%,P=0.028), resection rate (98.8%vs 88.7%,P<0.001) and less recurrence (2.9%vs 16.0%,P<0.001). Five-year overall survival was 92% atfer VATS and 92% atfer open thymectomy, with no signiifcant difference between the two groups (P=0.15). However, 5-year disease free survival were 92% in VATS group and 83% in Open group (P=0.011).Cox proportional hazards model revealed that WHO classiifcation, Masaoka-Koga stage and adjuvant therapy were independent predictive factors for overall survival, while surgical approach had no signiifcant impact on long-term outcome.ConclusionhTis study suggests that VATS thymectomy is an effective approach for clinically early-stage thymic malig-nancies. And it may offer better perioperative outcomes, as well as equal oncological survival.

Background and objectiveTo evaluate the surgical outcomes of tumor resection with or without total thymectomy for thymic epithelial tumors (TETs) using the Chinese Alliance for Research in Thymomas (ChART) retrospec-tive database.Methods Patients without preoperative therapy, who underwent surgery for early-stage (Masaoka-Koga stage I and II) tumors, were enrolled for the study. They were divided into thymectomy and thymomectomy groups according to the resection extent of the thymus. Demographic and surgical outcomes were compared between the two patients groups. Results A total of 1,047 patients were enrolled, with 796 cases in the thymectomy group and 251 cases in the thymomec-tomy group. Improvement rate of myasthenia gravis (MG) was higher atfer thymectomy than atfer thymomectomy (91.6%vs 50.0%,P<0.001). Ten-year overall survival was similar between the two groups (90.9% atfer thymectomy and 89.4% atfer thymomectomy,P=0.732). Overall, recurrence rate was 3.1% atfer thymectomy and 5.4% atfer thymomectomy, with no sig-niifcant difference between the two groups (P=0.149). Stratiifed analysis revealed no signiifcant difference in recurrence rates in Masaoka-Koga stage I tumors (3.2%vs 1.4%,P=0.259). However in patients with Masaoka-Koga stage II tumors, recurrence was signiifcantly less atfer thymectomy group than atfer thymomectomy (2.9%vs 14.5%,P=0.001).Conclusion hTymectomy, instead of tumor resection alone, should still be recommended as the surgical standard for thymic malignancies, especially for stage II tumors and those with concomitant MG.

Background and objectivePostoperative radiotherapy (PORT) for thymic tumor is still controversial. The object of the study is to evaluate the role of PORT for stage I/II/III thymic tumor.MethodsThe database of Chinese Al-liance of Research for Thymomas (ChART) was retrieved for patients with stage I/II/III thymic tumor who underwent surgi-cal therapy without neoajuvant therapy between 1994 and 2012. Univariate and multivariate survival analyses were performed. Cox proportional hazard model was used to determine the hazard ratio for death.Results 1,546 stage I/II/III patients were identiifed from ChART database. Among these patients, 649 (41.98%) underwent PORT. PORT was associated with gender, histologic type (World Health Organization, WHO), surgical extent, complete resection, Masaoka stage and adjuvant che-motherapy. The 5-yr and 10-yr overall survival (OS) rates and disease-free survival (DFS) rate for patients underwent surgery followed by PORT were 90% and 80%, 81% and 63%, comparing with 96% and 95%, 92% and 90% for patients underwent surgery alone (P=0.001,P<0.001) respectively. In univariate analysis, age, histologic type (WHO), Masaoka stage, complete-ness of resection, and PORT were associated with OS. Multivariable analysis showed that histologic type (WHO)(P=0.001), Masaoka stage (P=0.029) and completeness of resection (P=0.003) were independently prognostic factors of OS. In univari-ate analysis, gender, myasthenia gravis, histologic type (WHO), Masaoka stage, surgical approach, PORT and completeness of resection were associated with DFS. Multivariable analysis showed that histologic type (WHO) (P<0.001), Masaoka stage (P=0.005) and completeness of resection (P=0.006) were independently prognostic factors of DFS. Subgroup analysis showed that patients with incomplete resection underwent PORT achieved the better OS and DFS (P=0.010, 0.017, respectively). However, patients with complete resection underwent PORT had the worse OS and DFS (P<0.001,P<0.001, respectively). ConclusionThe current retrospective study indicated that PORT atfer incomplete resection could improve OS and DFS for patients with stage I/II/III thymic tumor. But for those atfer complete resection, PORT may not help improve prognosis on the whole.

Background and objectiveTo study the role of postoperative chemotherapy and its prognostic effect in Masaoka-Koga stage III and IV thymic tumors.Methods Between 1994 and 2012, 1,700 patients with thymic tumors who underwent surgery without neoajuvant therapy were enrolled for the study. Among them, 665 patients in Masaoka-Koga stage III and IV were further analyzed to evaluate the clinical value of postoperative chemotherapy. TheKaplan-Meier method was used to obtain the survival curve of the patients divided into different subgroups, and theCox regression analysis was used to make multivariate analysis on the factors affecting prognosis. A Propensity-Matched Study was used to evaluate the clinical value of chemotherapy.Results Two-hundred-twenty-one patients were treated with postoperative chemotherapy, while the rest 444 cases were not. The two groups showed signiifcant differences (P<0.05) regarding the incidence of myasthenia gravis, World Health Organization (WHO) histological subtypes, pathological staging, resection status and the use of postopera-tive radiotherapy. WHO type C tumors, incomplete resection, and postoperative radiotherapy were signiifcantly related to increased recurrence and worse survival (P<0.05). Five-year and 10-year disease free survivals (DFS) and recurrence rates in patients who underwent surgery followed by postoperative chemotherapy were 51% and 30%, 46% and 68%, comparing with 73% and 58%, 26% and 40% in patients who had no adjuvant chemotherapy atfer surgery (P=0.001,P=0.001, respectively). In propensity-matched study, 158 pairs of patients with or without postoperative chemotherapy (316 patients in total) were se-lected and compared accordingly. Similar 5-year survival rates were detected between the two groups (P=0.332). Conclusion Pathologically higher grade histology, incomplete resection, and postoperative radiotherapy were found to be associated with worse outcomes in advanced stage thymic tumors. At present, there is no evidence to show that postoperative chemotherapy may help improve prognosis in patients with Masaoka-Koga-Koga stage III and IV thymic tumors.