Error-Triggered Adaptive Sparse Identification for Predictive Control and Its Application to Multiple Operating Conditions Processes.

With the digital transformation of process manufacturing, identifying the system model from process data and then applying to predictive control has become the most dominant approach in process control. However, the controlled plant often operates under changing operating conditions. What is more, there are often unknown operating conditions such as first appearance operating conditions, which make traditional predictive control methods based on identified model difficult to adapt to changing operating conditions. Moreover, the control accuracy is low during operating condition switching. To solve these problems, this article proposes an error-triggered adaptive sparse identification for predictive control (ETASI4PC) method. Specifically, an initial model is established based on sparse identification. Then, a prediction error-triggered mechanism is proposed to monitor operating condition changes in real time. Next, the previously identified model is updated with the fewest modifications by identifying parameter change, structural change, and combination of changes in the dynamical equations, thus achieving precise control to multiple operating conditions. Considering the problem of low control accuracy during the operating condition switching, a novel elastic feedback correction strategy is proposed to significantly improve the control accuracy in the transition period and ensure accurate control under full operating conditions. To verify the superiority of the proposed method, a numerical simulation case and a continuous stirred tank reactor (CSTR) case are designed. Compared with some state-of-the-art methods, the proposed method can rapidly adapt to frequent changes in operating conditions, and it can achieve real-time control effects even for unknown operating conditions such as first appearance operating conditions.

EaLDL: Element-Aware Lifelong Dictionary Learning for Multimode Process Monitoring.

With the rapid development of modern industry and the increasing prominence of artificial intelligence, data-driven process monitoring methods have gained significant popularity in industrial systems. Traditional static monitoring models struggle to represent the new modes that arise in industrial production processes due to changes in production environments and operating conditions. Retraining these models to address the changes often leads to high computational complexity. To address this issue, we propose a multimode process monitoring method based on element-aware lifelong dictionary learning (EaLDL). This method initially treats dictionary elements as fundamental units and measures the global importance of dictionary elements from the perspective of the multimode global learning process. Subsequently, to ensure that the dictionary can represent new modes without losing the representation capability of historical modes during the updating process, we construct a novel surrogate loss to impose constraints on the update of dictionary elements. This constraint enables the continuous updating of the dictionary learning (DL) method to accommodate new modes without compromising the representation of previous modes. Finally, to evaluate the effectiveness of the proposed method, we perform comprehensive experiments on numerical simulations as well as an industrial process. A comparison is made with several advanced process monitoring methods to assess its performance. Experimental results demonstrate that our proposed method achieves a favorable balance between learning new modes and retaining the memory of historical modes. Moreover, the proposed method exhibits insensitivity to initial points, delivering satisfactory results under various initial conditions.

Integrated analysis of Mendelian Randomization and Bayesian colocalization reveals bidirectional causal association between inflammatory bowel disease and psoriasis.

BACKGROUND: Observational studies have suggested an association between inflammatory bowel disease [IBD] and psoriasis. However, the detailed genetic basis, causality, and direction of this association remain unclear. METHODS: Bidirectional two-sample Mendelian Randomization [MR] analysis was conducted using summary statistics from published genome-wide association studies. Bayesian Colocalization and multivariable MR [MVMR] analyses were performed to identify candidate variants and risk genes involved in the shared genetic basis between IBD, psoriasis, and their subtypes. RESULTS: Genetically predicted IBD and Crohn's disease [CD] were associated with an increased risk of psoriasis, psoriasis vulgaris [PsV], and psoriatic arthritis [PsA] (IBD on psoriasis: pooled odds ratio [OR] 1.09, 95% confidence interval [CI] 1.04-1.14, p = .0001; CD on psoriasis: pooled OR 1.10, 95% CI 1.06-1.15, p < .0001) and vice versa (psoriasis on IBD: pooled OR 1.11, 95%CI 1.02-1.21), whereas CD only exhibited a unidirectional association with psoriasis. Colocalization analysis revealed eight candidate genetic variants and risk genes (including LINC00824, CDKAL1, IL10, IL23R, DNAJC27, LPP, RUNX3, and RGS14) associated with a shared genetic basis. Among these, IL23R, DNAJC27, LPP, and RGS14 were further validated by MVMR analysis. CONCLUSION: Our findings indicated bidirectional causal associations between IBD and psoriasis (including PsV and PsA), which were attributed primarily to CD rather than Ulcerative colitis [UC]. Furthermore, we identified several candidate variants and risk genes involved in the shared genetic basis of IBD and psoriasis. Acquiring a better understanding of the shared genetic architecture underlying IBD and psoriasis would help improve clinical strategies.

Novel insights into the progression and prognosis of the calpain family members in hepatocellular carcinoma: a comprehensive integrated analysis.

Objectives: The goal of our bioinformatics study was to comprehensively analyze the association between the whole calpain family members and the progression and prognosis of hepatocellular carcinoma (HCC). Methods: The data were collected from The Cancer Genome Atlas (TCGA). The landscape of the gene expression, copy number variation (CNV), mutation, and DNA methylation of calpain members were analyzed. Clustering analysis was performed to stratify the calpain-related groups. The least absolute shrinkage and selection operator (LASSO)-based Cox model was used to select hub survival genes. Results: We found 14 out of 16 calpain members expressed differently between tumor and normal tissues of HCC. The clustering analyses revealed high- and low-risk calpain groups which had prognostic difference. We found the high-risk calpain group had higher B cell infiltration and higher expression of immune checkpoint genes HAVCR2, PDCD1, and TIGHT. The CMap analysis found that the histone deacetylase (HDAC) inhibitor trichostatin A and the PI3K-AKT-mTOR pathway inhibitors LY-294002 and wortmannin might have a therapeutic effect on the high-risk calpain group. The DEGs between calpain groups were identified. Subsequent univariate Cox analysis of each DEG and LASSO-based Cox model obtained a calpain-related prognostic signature. The risk score model of this signature showed good ability to predict the overall survival of HCC patients in TCGA datasets and external validation datasets from the Gene Expression Omnibus database and the International Cancer Genome Consortium database. Conclusion: We found that calpain family members were associated with the progression, prognosis, and drug response of HCC. Our results require further studies to confirm.

Autocorrelation Feature Analysis for Dynamic Process Monitoring of Thermal Power Plants.

Accurate process monitoring plays a crucial role in thermal power plants since it constitutes large-scale industrial equipment and its production safety is of great significance. Therefore, accurate process monitoring is very important for thermal power plants. The vigorous nature of the production process requires dynamic algorithms for monitoring. Since the common dynamic algorithm is mainly based on data expansion, the online computing complexity is too high because of data redundancy. Accordingly, this article proposes an innovative, dynamic process monitoring algorithm called autocorrelation feature analysis (AFA). AFA mines the dynamic information of continuous samples by calculating the correlation between the current time and past time features. While improving the monitoring effect, the AFA algorithm also has extremely low online computational complexity, even lower than common static algorithms, such as principal component analysis. Furthermore, this study exhibits the general form of dynamic additive faults for the first time and verifies the reliability of the algorithm through fault detectability analysis. Conclusively, the superiority of the AFA algorithm is verified on a numerical example, continuous stirred tank reactor (CSTR), and real data measured from a 1000-MW ultrasupercritical thermal power plant.

LSTMED: An uneven dynamic process monitoring method based on LSTM and Autoencoder neural network.

Due to the complicated production mechanism in multivariate industrial processes, different dynamic features of variables raise challenges to traditional data-driven process monitoring methods which assume the process data is static or dynamically consistent. To tackle this issue, this paper proposes a novel process monitoring method based on the long short-term memory (LSTM) and Autoencoder neural network (called LSTMED) for multivariate process monitoring with uneven dynamic features. First, the LSTM units are arranged in the encoder-decoder form to construct an end-to-end model. Then, the constructed model is trained in an unsupervised manner to capture long-term time dependency within variables and dominant representation of high dimensional process data. Afterward, the kernel density estimation (KDE) method is performed to determine the control limit only based on the reconstruction error from historical normal data. Finally, effective online monitoring for uneven dynamic process can be achieved. The performance and advantage of the process monitoring method proposed are explained through typical cases, including the numerical simulation and Tennessee Eastman (TE) benchmark process, and comparative experimental analysis with state-of-the-art methods.

A transfer predictive control method based on inter-domain mapping learning with application to industrial roasting process.

As a critical variable in the roasting process, the roasting temperature has a significant influence on operating conditions. Model predictive control (MPC) provides a path to stabilize the roasting temperature. However, process data collected at different periods usually follow different distributions due to the fluctuation of feed composition for the roasting process, result in a model mismatch on online control. For this reason, a transfer predictive control method based on inter-domain mapping learning (IDML-MPC) is proposed. The proposed method first treat historical and online data as two domains. Then, a distribution mapping function from one domain to another domain is learned to make the distribution of the historical data follow that of the online data. Finally, an accurate online prediction model is built, roasting temperature control is achieved by minimizing the cost function with respect to the predicted value and the control input. The effectiveness of the proposed method is demonstrated by comparative experiments based on a numerical example and a simulation platform of the roasting process. Experimental results compared with some state-of-the-art methods show that it is necessary to take into account the distribution differences between historical data and online data when production conditions change. The IDML-MPC improved the control performance for the roasting temperature with an average 56.98% reduction in the root mean square error.

Primary tumor location (right versus left side of the colon) and resection affect the survival of patients with liver metastases from colonic neuroendocrine carcinoma: a population-based study.

BACKGROUND: Colonic neuroendocrine carcinomas (co-NECs) are heterogeneous and aggressive, especially with regard to metastasis. Whether co-NECs on the right and left sides of the colon have different characteristics from colon adenocarcinoma is unknown. METHODS: The co-NEC patients were selected from the 2010-2017 Surveillance, Epidemiology, and End Results Program (SEER) database. The right and left sides of the colon were separated by the splenic flexure. Coarsened exact matching (CEM) was performed to adjust for relevant factors before regression models were constructed. RESULTS: A total of 669 pathologically diagnosed co-NEC patients with sufficient baseline data were identified from the SEER database. A total of 80.72% of the patients had co-NEC that originated from the right side of the colon, and their mean overall survival (mOS) was similar to that of the patients with left-sided co-NECs (right versus left: 22.30 m versus 22.55 m). A total of 44.84% of the patients were diagnosed with liver metastasis (46.68% right side versus 37.98% left side). In patients with liver metastasis, those with right-sided co-NECs had better survival than those with left-sided co-NECs (mOS right versus left: 15.37 m versus 9.62 m; adjusted hazard ratio (HR) = 0.69, 95% confidence interval (CI): 0.49-0.98, p = 0.035). To further investigate the survival benefits of primary site resection, we separated the patients who had liver metastasis according to the primary site and performed CEM to balance the groups (no patients underwent liver metastasis resection or intervention). The results suggested that primary surgery could benefit patients with both left- and right-sided co-NECs (adjusted HR = 0.50, 95% CI: 0.33-0.77, p = 0.001 on the right side; HR = 0.38, 95% CI: 0.16-0.89, p = 0.026 on the left side). CONCLUSIONS: Co-NECs frequently originate on the right side and commonly develop liver metastasis. Right-sided co-NECs are associated with better survival than left-sided co-NECs after liver metastasis has occurred. Primary site resection is associated with prolonged survival in co-NEC patients with liver metastasis, regardless of the side from which the co-NEC has originated.

SPARCL1 exhibits different expressions in left- and right-sided colon cancer and is downregulated via DNA methylation.

Aim: The authors previously found that SPARCL1 functions to suppress colorectal cancer metastasis. Here, the epigenetic mechanism of SPARCL1 regulation and its relationship with clinicopathological features in colon cancer were investigated. Materials & methods: SPARCL1 expression was evaluated by immunohistochemistry staining in a tissue array containing 271 left-sided colon cancer samples and 257 right-sided colon cancer samples. In vivo and in vitro DNA methylation states were measured by biochemical sulfide potential assay. The transcription and DNA methylation states in cells were altered by siRNA or decitabine treatment, respectively. Cellular motility properties were compared through transwell assay. Results & conclusion: SPARCL1, mediated by its DNA methylation, may arrest colorectal carcinoma motility. Furthermore, SPARCL1 expression is higher and may have a specific prognostic value in left-sided colon cancer.

Lay abstract The heterogeneity of colorectal cancer is largely based on their primary tumor locations. For example, patients with left-sided colon cancer generally have better prognoses and different treatment strategies in comparison with patients with right-sided colon cancer. Research is required to uncover the mechanisms underlying these differences. The authors showed that the addition of methyl groups (DNA methylation) to DNA encoding SPARCL1 (a protein involved in cell adhesion) could inhibit regular colorectal cell motility. Furthermore, SPARCL1 protein expression was higher and positively correlated with better prognosis specifically in patients with left-sided colon cancer. The present study provides an evidential basis for the different molecular biological features between left- and right-sided colon cancer. SPARCL1 is a promising candidate for predicting colon cancer prognosis and potential therapeutic intervention. Further in-depth study of SPARCL1 would be of great value for clinical application.

Pan-cancer analyses demonstrate that ANKRD6 is associated with a poor prognosis and correlates with M2 macrophage infiltration in colon cancer.

OBJECTIVE: Ankyrin repeat domain-containing protein 6 (ANKRD6) is an ankyrin repeat-containing protein which is structurally related to vertebrate inversin and Drosophila Diego. However, the correlations between ANKRD6 and tumor-infiltrating immune cells in cancers is not clear. METHODS: ANKRD6 expression was analyzed by Oncomine, Tumor Immune Estimation Resource (TIMER) and Gene Expression Profiling Interactive Analysis (GEPIA). PrognoScan and GEPIA were used to evaluate the influence of ANKRD6 on clinical prognosis. TIMER and CIBERSORT were used to analyze correlations between ANKRD6 expression levels and tumor immune cell infiltrates. Immunohistochemical analysis of the relationship between ANKRD6 expression and overall survival, as well as the relationship between ANKRD6 expression and M2 macrophage infiltration, was performed. RESULTS: High level of ANKRD6 expression was associated with poor prognosis of colon cancer. ANKRD6 expression level was positively correlated with infiltrating levels of CD8+ T cells, CD4+ T cells, macrophages, neutrophils and dendritic cells in colon cancer by using TIMER. Using CIBERSORT, we found that in plasma cells, CD8+ T cells, CD4+ memory resting T cells, follicular helper T cells and activated natural killer cells were significantly lower in the ANKRD6-high group than in the ANKRD6-low group. M0 and M2 macrophages were significantly higher in the ANKRD6-high group than in the ANKRD6-low group. Immunohistochemistry confirmed that M2 macrophage infiltration in the ANKRD6-high group significantly increased. CONCLUSIONS: The high ANKRD6 expression is associated with poor prognosis of colon cancer. ANKRD6 expression is positively correlated with M2 macrophage infiltration in colon cancer.

Factors Affecting the Outcomes of Patients with Malignant Rhabdoid Tumors: A Population-Based Study.

Objective: Malignant rhabdoid tumor (MRT) is a rare but aggressive malignancy. It has been a long time since data on this tumor have been updated. Methods: We retrospectively reviewed patients from the SEER database who were pathologically diagnosed with MRT and analyzed incidence rates, clinical features and survival using Stata 12.0. Results: In total, 544 patients were included in the epidemiological analysis. There were two peak periods of MRT incidence: patients younger than 4 years and those older than 70 years. Further survival analysis showed that the survival of children (especially younger than 1 year) was markedly worse than that of adults (P<0.01), and different primary sites were associated with different age groups and survival outcomes. The central nervous system (CNS) was the most common primary site (50.00%), followed by the kidney (15.66%). Patients with MRTs that originated from the digestive system experienced worse survival outcomes than those with MRTs originating from other locations. Primary site surgery conferred survival benefits to patients with renal and digestive system MRTs (HR = 0.06, CI: 0.02-0.23, P<0.01; HR=0.10, CI: 0.02-0.48, P<0.01), whereas radiotherapy conferred benefits to patients with CNS, bone and soft tissue MRTs (HR=0.22, CI: 0.15-0.34, P<0.01; HR=0.44, CI: 0.21-0.90 P=0.03). Conclusions: Our results indicate that age and the primary site of MRT are critical clinical factors that affect patient survival and treatment choices. Primary site tumor resection should be considered for renal and digestive system MRTs, and systematic therapy, including surgery and radiotherapy, should be recommended for the treatment of CNS, bone and soft tissue MRTs.

Immunotherapies for well-differentiated grade 3 gastroenteropancreatic neuroendocrine tumors: A new category in the World Health Organization classification.

According to the 2019 World Health Organization (WHO) classification, well-differentiated grade 3 (G3) gastroenteropancreatic (GEP) neuroendocrine tumors (NETs) are a new category of cancer of the digestive system. G3 GEP-NET research and treatment are not as robust as those of lower grade (G1/2) NETs and poorly differentiated neuroendocrine carcinomas (NECs). Previously, the management of high-grade NETs was mainly based on NEC therapies, as high-grade NETs were classified as NECs under the previous WHO classification. Despite this, G3 GEP-NETs are significantly less responsive to platinum-based chemotherapy regimens than NECs, due to their distinct molecular pathogenesis and course of pathological grade transition. Patients with advanced G3 GEP-NETs, who have progressed or are intolerant to chemotherapy regimens such as capecitabine plus temozolomide, have limited treatment choices. Immunotherapy has helped patients with a variety of cancers attain long-term survival through the use of immune checkpoint inhibitors. Immunotherapies, either alone or in combination with other therapies, do not have a clear function in the treatment of G3 GEP-NETs. Currently, the majority of immunotherapy studies, both prospective and retrospective, do not reliably differentiate G3 GEP-NETs from NECs. By contrast, a significant number of studies include non-GEP neuroendocrine neoplasms (NENs). Therefore, there is an urgent need to summarize and evaluate these data to provide more effective therapeutic approaches for patients with this rare tumor. The purpose of this mini-review was to screen and summarize information on G3 GEP-NETs from all studies on NENs immunotherapy.

Ultra-mutated colorectal cancer patients with POLE driver mutations exhibit distinct clinical patterns.

POLE mutations, which lead to an ultramutated phenotype in colorectal cancer (CRC), have been reported as a promising marker in immunotherapy. We performed sequencing of CRC cases in Zhejiang University (ZJU) and extracted obtainable data from recently published results, including The Cancer Genome Atlas (TCGA), Japanese studies and clinical trials, to present clinical patterns of POLE driver-mutated CRC and reveal its heterogeneity. The rate of somatic POLE driver mutations has been reported as 2.60% (ZJU cohort), 1.50% (TCGA cohort), 1.00% (Japan cohort), and 1.00% (Lancet cohort). POLE driver mutations show a clearly increased mutation burden (mean TMB: 217.98 mut/Mb in ZJU; 203.13 mut/Mb in TCGA). Based on pooled data, more than 70.00% of patients with POLE driver mutations were diagnosed before they were 55 years old and at an early disease stage (Stage 0-II >70.00%), and more than 70.00% were male. Among Asian patients, 68.40% developed POLE driver mutations in the left-side colon, whereas 64.00% of non-Asian patients developed them in the right-side colon (p < 0.01). The top three amino acid changes due to POLE driver mutations are P286R, V411L, and S459F. Investigators and physicians should ascertain the heterogeneity and clinical patterns of POLE driver mutations to be better equipped to design clinical trials and analyze the data.

A Retrospective Exploration of Targeted Maintenance Therapy in Advanced Colorectal Cancer: Based on the Background of Chinese Patient Assistance Program.

Background: Maintenance therapy with bevacizumab (Bev) in patients with colorectal cancer (CRC) provides progression-free survival (PFS) benefits. However, the role of maintenance therapy with an anti-EGFR monoclonal antibody has not been established. Methods: Eligible CRC patients were assigned to maintenance therapy with cetuximab (Cet; Cet group) or Bev (Bev group). PFS, the duration of maintenance therapy, and safety were analyzed. Cox multivariate regression analyses were performed to determine independent prognostic factors. Results: A total of 143 eligible patients were assigned to the Cet (n = 79) or Bev (n = 64) groups. In the Cet group, all patients had KRAS wild-type. The baseline characteristics were well-balanced between the two groups, except for a higher percentage of patients with a left-sided primary tumor in the Cet group than in the Bev group (86.1 vs. 62.5%, P < 0.0001). The median PFS was not significantly different between the Cet group and the Bev group: 5.9 months (95% CI 2.30-9.50) vs. 7.0 months (95% CI 3.69-10.31) (HR 1.17, 95% CI 0.77-1.79, P = 0.45). The median duration of maintenance therapy in the Cet group was shorter than that in the Bev group: 4.0 months (95% CI 1.94-5.99) vs. 4.8 months (95% CI 2.68-6.98) (HR 0.90, 95% CI 0.61-1.33; P = 0.59). The subgroup analyses showed that the median PFS for the first maintenance therapy and the second maintenance therapy were 3.2 months (95% CI 1.69-4.78) and 5.2 months (95% CI 1.58-8.83), respectively (HR 0.89, 95% CI 0.44-1.81; P = 0.75). Conclusions: This study suggests that maintenance therapy with Cet or Bev can be considered an appropriate option following induction chemotherapy for selected patients with advanced CRC. Multiple maintenance therapy seems to confer survival benefits in advanced CRC. Maintenance therapy with Cet after first-line induction chemotherapy seems to be associated with greater survival benefits.

A 10-year Population-based Study of the Differences between NECs and Carcinomas of the Esophagus in Terms of Clinicopathology and Survival.

Purpose: The prevalence of esophageal NECs is rising, but to date, no studies have compared its clinicopathological characteristics to those of esophageal ACs and SCCs from the same period. Patients and methods: A 10-year population-based retrospective cohort study was conducted with the Surveillance, Epidemiology, and End Results Program database. Statistical analyses were performed using Intercooled Stata 12.0 software. Results: A total of 17,196 eligible patients with esophageal tumors, including 246 NECs, 6,102 SCCs and 10,848 ACs, were analyzed. ACs showed an obviously higher prevalence than the other two tumor types, and the prevalence of NECs was increasing. NECs were associated with an obviously worse survival than ACs (log-rank test, P<0.01). Most NECs were poorly differentiated and had an obviously higher percentage of metastasis. NECs and ACs often metastasized to the liver (29.41% and 23.11%, respectively), while SCCs typically metastasized to the lung (15.84%) and distant lymph nodes (15.37%). We divided the patients into two groups for further analysis according to the metastasis status. For NECs, no benefit was obtained by surgery in metastatic disease. For SCCs and ACs, surgery of the primary sites produced survival benefits in both groups, but the benefits of lymphadenectomy and metastasis dissection need further study. Conclusion: NECs of the esophagus have the worst prognosis compared to SCCs and ACs from the same period. Radical surgery provides limited benefits to patients diagnosed with NECs, so systemic treatments should be considered instead of surgical procedures. A unique guideline with a new staging and grading system for esophageal NECs is urgently needed.

A novel 4-gene prognostic signature for hypermutated colorectal cancer.

BACKGROUND: Hypermutated colorectal cancer (CRC) reportedly accounts for 15%-17% of all cases of CRC. However, the proportion and number of patients with hypermutated CRC cannot be unappreciated. Additionally, therapy options for these patients differ from those for CRC patients, with a greater potential benefit from immunotherapy. MATERIALS AND METHODS: We sequenced the tumor mucosa of CRC patients with >24 months of follow-up data at our center and identified mutation profiles of hypermutated CRC as a training data set (Zhejiang University [ZJU]); we then collected patients from The Cancer Genome Atlas (TCGA) as a validation data set. Recurrently mutated genes were combined to calculate a compound score via Cox proportional hazards model. Patients with higher-than-median scores were segregated as the high-risk group. Outcomes were analyzed by Kaplan-Meier and Cox regression analyses using Python (3.6.0) and R (3.4.0). RESULTS: We constructed a 4-gene signature (ACVR2A, APC, DOCK2, and POLE), with training in 45 hypermutated patients at ZJU and validation in 24 hypermutated patients from TCGA. Patients in the high-risk group showed poor survival (adjusted HR =9.85, 95% CI: 2.07-46.81, P=0.004). Further subgroup analysis was performed for stage II and III colon cancer (HR =10.91, 95% CI: 1.36-87.5, P=0.005) and high microsatellite instability (MSI-H) CRC (HR =12.57, 95% CI: 1.57-100.69, P=0.002) subgroups, which verified that our signature is universal. We then compared our prognostic signature with other risk factors (including MSI status, POLE driver mutation, BRAF-p.V600E, tumor mutational burden, and TNM staging). The results proved that our 4-gene signature is better than the other risk factor for prognosis in hypermutated CRC. CONCLUSION: Our 4-gene signature is a good predictor of survival for hypermutated CRC, and this signature is powerful in stage II and III colon cancer and MSI-H CRC. Future prospective studies are needed to confirm the power of the 4-gene signature in patients receiving immunotherapy.

EGFR Target Therapy Combined with Gemox for Advanced Biliary Tract Cancers: a Meta-analysis based on RCTs.

Background: Controversy exists regarding whether EGFR-targeted therapy combined with GEMOX (gemcitabine and oxaliplatin) provides additional benefits over GEMOX alone for biliary tract cancer patients. Therefore, this meta-analysis of randomized controlled trials (RCTs) was performed to assess the efficacy and safety of the GEMOX + EGFR-targeted regimen, and subgroup analysis was conducted to identify groups that might benefit from targeted therapy. Methods: The PubMed, Cochrane Library, and ClinicalTrials.gov registries were searched for published studies. Hazard ratios (HRs) for progression-free survival (PFS) and overall survival (OS) were pooled using a fix-effect model. Risk-ratios (RRs) were used to analyse the objective response rate (ORR) and adverse events. Results: Four RCTs were assessed. GEMOX + EGFR-targeted therapy significantly improved PFS (HR = 0.80, 95% CI 0.66-0.94, P = 0.03) and was associated with a better ORR (RR = 1.52, 95% CI 1.13-2.04, P = <0.01), whereas the TKI group achieved a better ORR in subgroup analysis. Patients with cholangiocarcinoma responded well to the GEMOX + EGFR-targeted regimen, leading to a better ORR (RR = 1.78, 95% CI 1.21-2.61, P = <0.01). Unfortunately, PFS benefits were not translated into OS benefits (HR = 0.92, 95% CI 0.75-1.08, P = 0.39). Conclusion: GEMOX + EGFR-targeted therapy is a considerable and tolerable treatment option for patients with advanced BTCs, improving both PFS and ORR but not prolonging patient survival. Patients with cholangiocarcinoma would benefit the most from EGFR-targeted therapy.