Zebin Xiao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Zebin Xiao has a total of 309 co-authors. Here are the top co-authors by number of shared publications:
- Ellen Puré
Department of Biomedical Sciences, University of Pennsylvania, Philadelphia, USA
42 shared publications - Carl H. June
University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania, United States
38 shared publications - Leslie Todd
University of Pennsylvania 1 Department of Biomedical Sciences Philadelphia, Pennsylvania
23 shared publications - Steven M. Albelda
Center for Cellular Immunology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA
22 shared publications - Zuohua Tang
Department of Radiology, Eye & ENT Hospital, Fudan University, Shanghai, China
18 shared publications - Wei Guo
Department of Biology, School of Arts and Sciences, University of Pennsylvania, Philadelphia, USA
17 shared publications
Latest Publications
- Open Access
- 8 September 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1)
Deep learning CT signature for predicting early liver metastases in pancreatic ductal adenocarcinoma
Accurately predicting the risk of early liver metastases (ELM) and identifying patients who are most likely to benefit from neoadjuvant therapy (NAT) are critical for pancreatic ductal adenocarcinoma (PDAC). Here, we develop a Mamba-based predictive model that integrates imaging features from both the primary pancreatic tumor and the liver to assess the risk of ELM. The model is evaluated in a multi-institutional cohort of 1063 PDAC patients and demonstrates robust performance in predicting ELM (AUCs: 0.806-0.890). Besides, model-defined high-risk patients exhibit significantly shorter progression-free survival (PFS: HR = 1.93, p < 0.001) and overall survival (OS: HR = 1.89, p < 0.001). Notably, NAT confers significant OS benefits in model-defined high-risk patients (17.4 vs. 34.1 months; p < 0.001), even after propensity score matching (p = 0.004), while no survival benefit occurs in low-risk patients. Radiotranscriptomic analyses further reveal relative biological aggressiveness in the high-risk group. Overall, our proposed Mamba-based framework enables accurate prediction of ELM and may serve as a clinically actionable tool for identifying PDAC patients most likely to benefit from NAT.
- 8 September 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1)
Open AccessDeep learning CT signature for predicting early liver metastases in pancreatic ductal adenocarcinoma
Accurately predicting the risk of early liver metastases (ELM) and identifying patients who are most likely to benefit from neoadjuvant therapy (NAT) are critical for pancreatic ductal adenocarcinoma (PDAC). Here, we develop a Mamba-based predictive model that integrates imaging features from both the primary pancreatic tumor and the liver to assess the risk of ELM. The model is evaluated in a multi-institutional cohort of 1063 PDAC patients and demonstrates robust performance in predicting ELM (AUCs: 0.806-0.890). Besides, model-defined high-risk patients exhibit significantly shorter progression-free survival (PFS: HR = 1.93, p < 0.001) and overall survival (OS: HR = 1.89, p < 0.001). Notably, NAT confers significant OS benefits in model-defined high-risk patients (17.4 vs. 34.1 months; p < 0.001), even after propensity score matching (p = 0.004), while no survival benefit occurs in low-risk patients. Radiotranscriptomic analyses further reveal relative biological aggressiveness in the high-risk group. Overall, our proposed Mamba-based framework enables accurate prediction of ELM and may serve as a clinically actionable tool for identifying PDAC patients most likely to benefit from NAT.
- Open Access
- 17 June 2026
- Published by American Association for Cancer Research (AACR)
Data from Predicting c-KIT Inhibitor Efficacy in Patient-Derived Models of Sinonasal Mucosal Melanomas through Integrated Histogram Analysis of Whole-Tumor DKI, IVIM, and DCE-MRI
Purpose:To evaluate whole-tumor histogram analysis of diffusion kurtosis imaging (DKI), intravoxel incoherent motion (IVIM), and dynamic contrast-enhanced MRI (DCE-MRI) in predicting the efficacy of imatinib, a c-KIT inhibitor, for treating patient-derived models derived from sinonasal mucosal melanomas (MM).Experimental Design:This study included 38 patients with histologically confirmed sinonasal MM, who underwent DKI, IVIM, and DCE-MRI. Patient-derived tumor xenograft models and precision-cut tumor slices were established to evaluate tumor response to imatinib. Whole-tumor histogram analysis was conducted on imaging parameters, and logistic regression models were applied to determine the predictive value of these metrics in differentiating responders from nonresponders.Results:Among the 38 patients with sinonasal MM, 12 were classified as responders and 26 as nonresponders based on patient-derived tumor xenograft and precision-cut tumor slice model responses to imatinib. The DKI model revealed significant differences in mean, median, 10th percentile, and 90th percentile values of Dk and K between responders and nonresponders (P < 0.05). The IVIM model indicated significant differences in 10th percentile and mean values of D, with kurtosis f being a strong predictor. The DCE-MRI model, using the 90th percentile Ktrans metric, demonstrated robust predictive performance, achieving an AUC of 0.89, with 80.77% specificity and 91.67% sensitivity. The combined logistic model integrating DKI, IVIM, and DCE-MRI metrics produced the highest predictive accuracy, with an AUC of 0.90.Conclusions:Whole-tumor histogram analysis of DKI, IVIM, and DCE-MRI offers a noninvasive method for predicting the efficacy of c-KIT inhibitors in sinonasal MMs, presenting valuable implications for guiding targeted treatment in this rare cancer type.
- 17 June 2026
- Published by American Association for Cancer Research (AACR)
Open AccessData from Predicting c-KIT Inhibitor Efficacy in Patient-Derived Models of Sinonasal Mucosal Melanomas through Integrated Histogram Analysis of Whole-Tumor DKI, IVIM, and DCE-MRI
Purpose:To evaluate whole-tumor histogram analysis of diffusion kurtosis imaging (DKI), intravoxel incoherent motion (IVIM), and dynamic contrast-enhanced MRI (DCE-MRI) in predicting the efficacy of imatinib, a c-KIT inhibitor, for treating patient-derived models derived from sinonasal mucosal melanomas (MM).Experimental Design:This study included 38 patients with histologically confirmed sinonasal MM, who underwent DKI, IVIM, and DCE-MRI. Patient-derived tumor xenograft models and precision-cut tumor slices were established to evaluate tumor response to imatinib. Whole-tumor histogram analysis was conducted on imaging parameters, and logistic regression models were applied to determine the predictive value of these metrics in differentiating responders from nonresponders.Results:Among the 38 patients with sinonasal MM, 12 were classified as responders and 26 as nonresponders based on patient-derived tumor xenograft and precision-cut tumor slice model responses to imatinib. The DKI model revealed significant differences in mean, median, 10th percentile, and 90th percentile values of Dk and K between responders and nonresponders (P < 0.05). The IVIM model indicated significant differences in 10th percentile and mean values of D, with kurtosis f being a strong predictor. The DCE-MRI model, using the 90th percentile Ktrans metric, demonstrated robust predictive performance, achieving an AUC of 0.89, with 80.77% specificity and 91.67% sensitivity. The combined logistic model integrating DKI, IVIM, and DCE-MRI metrics produced the highest predictive accuracy, with an AUC of 0.90.Conclusions:Whole-tumor histogram analysis of DKI, IVIM, and DCE-MRI offers a noninvasive method for predicting the efficacy of c-KIT inhibitors in sinonasal MMs, presenting valuable implications for guiding targeted treatment in this rare cancer type.
- Open Access
- 2 April 2026
- Published by American Association for Cancer Research (AACR)
Data from Targeted Lipid Nanoparticle Delivery of FAP-CAR mRNA Enables Potent In Vivo T-cell Engineering against Pancreatic Tumors
Fibroblast activation protein (FAP), which is highly expressed on cancer-associated fibroblasts (CAF), is a promising therapeutic target to achieve normalization of the tumor microenvironment. We previously established an ex vivo retroviral-transduced FAP-specific chimeric antigen receptor (FAP-CAR) T-cell approach to deplete FAP+ CAFs that resulted in delayed tumor growth associated with disruption of desmoplastic matrix and enhanced immune cell infiltration and reversed immune exclusion and immunosuppression. In this study, we describe an in vivo strategy for generating FAP-CAR T cells using anti-CD5–conjugated targeted lipid nanoparticles (tLNP) encapsulating FAP-CAR mRNA and assessed the efficacy of this approach compared with adoptive transfer of retrovirus-transduced CAR T cells in a preclinical model of pancreatic ductal adenocarcinoma. With transient CAR expression in >45% of splenic, >69% of circulating, and >35% of tumor-infiltrating T cells, the abundance of peripheral and intratumoral FAP-CAR+ T cells detected following a single intravenous dose of FAP-CAR mRNA tLNPs was greater than that detected following administration of 1 × 107ex vivo retrovirally transduced FAP-CAR T cells. Furthermore, in vivo mRNA CAR T-cell engineering resulted in as good or greater inhibition of tumor growth as compared with adoptive transfer of ex vivo retroviral-engineered T cells. Given that in vivo generation of CAR T cells resulted in transient CAR expression and circumvented the need for autologous T-cell isolation, viral vectors, and lymphodepletion, this platform represents a potentially safer, more accessible, and cost-effective method for targeting stromal cells to normalize the tumor microenvironment in desmoplastic tumors and has potential implications for tumor antigen–targeted CAR T cells.
- 2 April 2026
- Published by American Association for Cancer Research (AACR)
Open AccessData from Targeted Lipid Nanoparticle Delivery of FAP-CAR mRNA Enables Potent In Vivo T-cell Engineering against Pancreatic Tumors
Fibroblast activation protein (FAP), which is highly expressed on cancer-associated fibroblasts (CAF), is a promising therapeutic target to achieve normalization of the tumor microenvironment. We previously established an ex vivo retroviral-transduced FAP-specific chimeric antigen receptor (FAP-CAR) T-cell approach to deplete FAP+ CAFs that resulted in delayed tumor growth associated with disruption of desmoplastic matrix and enhanced immune cell infiltration and reversed immune exclusion and immunosuppression. In this study, we describe an in vivo strategy for generating FAP-CAR T cells using anti-CD5–conjugated targeted lipid nanoparticles (tLNP) encapsulating FAP-CAR mRNA and assessed the efficacy of this approach compared with adoptive transfer of retrovirus-transduced CAR T cells in a preclinical model of pancreatic ductal adenocarcinoma. With transient CAR expression in >45% of splenic, >69% of circulating, and >35% of tumor-infiltrating T cells, the abundance of peripheral and intratumoral FAP-CAR+ T cells detected following a single intravenous dose of FAP-CAR mRNA tLNPs was greater than that detected following administration of 1 × 107ex vivo retrovirally transduced FAP-CAR T cells. Furthermore, in vivo mRNA CAR T-cell engineering resulted in as good or greater inhibition of tumor growth as compared with adoptive transfer of ex vivo retroviral-engineered T cells. Given that in vivo generation of CAR T cells resulted in transient CAR expression and circumvented the need for autologous T-cell isolation, viral vectors, and lymphodepletion, this platform represents a potentially safer, more accessible, and cost-effective method for targeting stromal cells to normalize the tumor microenvironment in desmoplastic tumors and has potential implications for tumor antigen–targeted CAR T cells.
- Open Access
- 2 April 2026
- Published by American Association for Cancer Research (AACR)
Supplementary Figure 3 from Targeted Lipid Nanoparticle Delivery of FAP-CAR mRNA Enables Potent In Vivo T-cell Engineering against Pancreatic Tumors
Supplementary Figure 3. Kinetics of CD5-tLNO-FAP-CAR expression in primary murine T cells in vitro 24-, 48-, and 72-hour post incubation.
- 2 April 2026
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Figure 3 from Targeted Lipid Nanoparticle Delivery of FAP-CAR mRNA Enables Potent In Vivo T-cell Engineering against Pancreatic Tumors
Supplementary Figure 3. Kinetics of CD5-tLNO-FAP-CAR expression in primary murine T cells in vitro 24-, 48-, and 72-hour post incubation.
- Open Access
- 2 April 2026
- Published by American Association for Cancer Research (AACR)
Supplementary Figure 5 from Targeted Lipid Nanoparticle Delivery of FAP-CAR mRNA Enables Potent In Vivo T-cell Engineering against Pancreatic Tumors
Supplementary Figure 5. Characterization of FAP-CAR+ immune and stromal populations following in vivo T cell engineering.
- 2 April 2026
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Figure 5 from Targeted Lipid Nanoparticle Delivery of FAP-CAR mRNA Enables Potent In Vivo T-cell Engineering against Pancreatic Tumors
Supplementary Figure 5. Characterization of FAP-CAR+ immune and stromal populations following in vivo T cell engineering.
- Open Access
- 2 April 2026
- Published by American Association for Cancer Research (AACR)
Supplementary Figure 1 from Targeted Lipid Nanoparticle Delivery of FAP-CAR mRNA Enables Potent In Vivo T-cell Engineering against Pancreatic Tumors
Supplementary Figure 1. Flow cytometry gating strategy.
- 2 April 2026
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Figure 1 from Targeted Lipid Nanoparticle Delivery of FAP-CAR mRNA Enables Potent In Vivo T-cell Engineering against Pancreatic Tumors
Supplementary Figure 1. Flow cytometry gating strategy.
- Open Access
- 2 April 2026
- Published by American Association for Cancer Research (AACR)
Supplementary Figure 6 from Targeted Lipid Nanoparticle Delivery of FAP-CAR mRNA Enables Potent In Vivo T-cell Engineering against Pancreatic Tumors
Supplementary Figure 6. CAR Expression in Resting T Cells and in Tumors at 10 days Post-Treatment.
- 2 April 2026
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Figure 6 from Targeted Lipid Nanoparticle Delivery of FAP-CAR mRNA Enables Potent In Vivo T-cell Engineering against Pancreatic Tumors
Supplementary Figure 6. CAR Expression in Resting T Cells and in Tumors at 10 days Post-Treatment.