Bohai Feng

Dr. Bohai Feng

orcid

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
Verified
Journal Article
Open access
PDF file
CWChristopher Weusthof
Christopher Weusthof
SBSebastian Burkart
Sebastian Burkart
KSKarl Semmelmayer
Karl Semmelmayer
FSFabian Stögbauer
Fabian Stögbauer
BFBohai Feng
Bohai Feng
JHJochen Hess
Jochen Hess
Perineural invasion is a prevalent pathological finding in head and neck squamous cell carcinoma and a risk factor for unfavorable survival. An adequate diagnosis of perineural invasion by pathologic examination is limited due to the availability of tumor samples from surgical resection, which can arise in cases of definitive nonsurgical treatment. To address this medical need, we established a random forest prediction model for the risk assessment of perineural invasion, including occult perineural invasion, and characterized distinct cellular and molecular features based on our new and extended classification. RNA sequencing data of head and neck squamous cell carcinoma from The Cancer Genome Atlas were used as a training cohort to identify differentially expressed genes that are associated with perineural invasion. A random forest classification model was established based on these differentially expressed genes and was validated by inspection of H&E-stained whole image slides. Differences in epigenetic regulation and the mutational landscape were detected by an integrative analysis of multiomics data and single-cell RNA-sequencing data were analyzed. We identified a 44-gene expression signature related to perineural invasion and enriched for genes mainly expressed in cancer cells according to single-cell RNA-sequencing data. A machine learning model was trained based on the expression pattern of the 44-gene set with the unique feature to predict occult perineural invasion. This extended classification model enabled a more accurate analysis of alterations in the mutational landscape and epigenetic regulation by DNA methylation as well as quantitative and qualitative differences in the cellular composition in the tumor microenvironment between head and neck squamous cell carcinoma with or without perineural invasion. In conclusion, the newly established model could not only complement histopathologic examination as an additional diagnostic tool but also guide the identification of new drug targets for therapeutic intervention in future clinical trials with head and neck squamous cell carcinoma patients at a higher risk for treatment failure due to perineural invasion.

8 March 2021 in Cancers
Verified
Journal Article
Open access
PDF file
BFBohai Feng
Bohai Feng
JHJochen Hess
Jochen Hess
Immunotherapy by immune checkpoint inhibition has become a main pillar in the armamentarium to treat head and neck cancer and is based on the premise that the host immune system can be reactivated to successfully eliminate cancer cells. However, the response rate remains low and only a small subset of head and neck cancer patients achieves a durable clinical benefit. The availability of multi-omics data and emerging computational technologies facilitate not only a deeper understanding of the cellular composition in the tumor immune microenvironment but also enables the study of molecular principles in the complex regulation of immune surveillance versus tolerance. These knowledges will pave the way to apply immunotherapy more precisely and effectively. This review aims to provide a holistic view on how the immune landscape dictates the tumor fate and vice versa, and how integrative analysis of multi-omics data contribute to our current knowledge on the accuracy of predictive biomarkers and on a broad range of factors influencing the response to immunotherapy in head and neck cancer.

28 February 2025 in npj Precision Oncology
Verified
Journal Article
Open access
PDF file
BFBohai Feng
Bohai Feng
DZDi Zhao
Di Zhao
ZZZheng Zhang
Zheng Zhang
RJRu Jia
Ru Jia
PSPatrick J. Schuler
Patrick J. Schuler
JHJochen Hess
Jochen Hess
Head and neck squamous cell carcinoma (HNSC) is a prevalent malignancy, with HPV-negative tumors exhibiting aggressive behavior and poor prognosis. Understanding the intricate interactions within the tumor microenvironment (TME) is crucial for improving prognostic models and identifying therapeutic targets. Using BulkSignalR, we identified ligand-receptor interactions in HPV-negative TCGA-HNSC cohort (n = 395). A prognostic model incorporating 14 ligand-receptor pairs was developed using random forest survival analysis and LASSO-penalized Cox regression based on overall survival and progression-free interval of HPV-negative tumors from TCGA-HNSC. Multi-omics analysis revealed distinct molecular features between risk groups, including differences in extracellular matrix remodeling, angiogenesis, immune infiltration, and APOBEC enzyme activity. Deep learning-based tissue morphology analysis on HE-stained whole slide images further improved risk stratification, with region selection via Silicon enhancing accuracy. The integration of routine histopathology with deep learning and multi-omics data offers a clinically accessible tool for precise risk stratification, facilitating personalized treatment strategies in HPV-negative HNSC.

15 October 2021 in Cancers
Verified
Journal Article
Open access
PDF file
BFBohai Feng
Bohai Feng
KWKai Wang
Kai Wang
EHEsther Herpel
Esther Herpel
MPMichaela Plath
Michaela Plath
WWWilko Weichert
Wilko Weichert
JHJochen Hess
Jochen Hess
Squamous cell carcinoma (SCC) is the most prevalent histological type of human cancer, including head and neck squamous cell carcinoma (HNSCC). However, reliable prognostic gene signatures for SCC and underlying genetic and/or epigenetic principles are still unclear. We identified 37 prognostic candidate genes by best cutoff computation based on survival in a pan-SCC cohort (n = 1334) of The Cancer Genome Atlas (TCGA), whose expression stratified not only the pan-SCC cohort but also independent HNSCC validation cohorts into three distinct prognostic subgroups. The most relevant prognostic genes were prioritized by a Least Absolute Shrinkage and Selection Operator Cox regression model and were used to identify subgroups with high or low risks for unfavorable survival. An integrative analysis of multi-omics data identified FN1, SEMA3A, CDH2, FBN1, COL5A1, and ADAM12 as key nodes in a regulatory network related to the prognostic phenotype. An in-silico drug screen predicted two MEK inhibitors (Trametinib and Selumetinib) as effective compounds for high-risk SCC based on the Cancer Cell Line Encyclopedia, which is supported by a higher p-MEK1/2 immunohistochemical staining of high-risk HNSCC. In conclusion, our data identified a molecular classifier for high-risk HNSCC as well as other SCC patients, who might benefit from treatment with MEK inhibitors.

26 March 2026 in Frontiers in Immunology
Verified
Journal Article
Open access
BFBohai Feng
Bohai Feng
YZYongwu Zhu
Yongwu Zhu
ZZZheng Zhang
Zheng Zhang
YWYu Wang
Yu Wang
PSPatrick J. Schuler
Patrick J. Schuler
JHJochen Hess
Jochen Hess
Background: Lactate metabolism is a hallmark of cancer metabolic reprogramming, shaping tumor immunity and therapeutic resistance, yet clinically accessible and low-cost methods to assess intratumoral lactate activity remain limited. Methods: We curated a lactate-related 59-gene signature and characterized its biological and clinical relevance across TCGA, GEO, and single-cell RNA-seq datasets. By integrating multi-omic, spatial, and computational analyses, we linked lactate metabolism to the tumor microenvironment and developed a deep learning framework to infer lactate metabolic states directly from routine H&E whole-slide images. Results: High lactate activity (LAC_H) was associated with enhanced tumor proliferation, suppressed immune infiltration, and poor response to both immunotherapy and radiotherapy in HNSCC. The pathology-based model achieved robust performance in distinguishing LAC_H from LAC_L tumors (AUC = 0.73–0.82 in HNSCC) and demonstrated strong generalizability across 12 TCGA cancer types (AUC = 0.78–0.89). Importantly, external validation in an independent real-world SAZHU-HNSCC cohort confirmed that model-predicted LAC_H tumors exhibited significantly increased protein expression of LDHA and MCT1 by immunohistochemistry, supporting the biological validity of the digital lactate biomarker. Conclusions: This study integrates multi-omics and digital pathology to infer tumor lactate metabolism from routine histology, providing a scalable and clinically practical digital biomarker for metabolism-informed precision oncology.

3 May 2026 in Cancer Medicine
Verified
Journal Article
Open access
XJXiaoli Jin
Xiaoli Jin
JXJing Xu
Jing Xu
YHYeting Hu
Yeting Hu
XCXin Chen
Xin Chen
SLSuyun Li
Suyun Li
BFBohai Feng
Bohai Feng
Background: Response to neoadjuvant chemoradiotherapy (NACRT) in locally advanced rectal cancer (LARC) is highly heterogeneous. Reliable pretreatment prediction of tumor regression and prognosis remains an unmet clinical need to optimize personalized management. Methods: A multicenter retrospective study of 434 Stage II‐III LARC patients was conducted across three tertiary hospitals. Pretreatment T2‐weighted MRI was used to build intratumoral and peritumoral macrohabitats. Local radiomic features within the tumor were clustered using K‐means to generate intratumoral habitats, with the optimal cluster number determined by the Calinski–Harabasz score. Radiomic and 3D deep‐learning features from each habitat and the peritumoral region were fused after LASSO‐based selection. Machine‐learning classifiers (support vector machine, logistic regression, multilayer perceptron) were trained to predict tumor regression grade (TRG). Performance was assessed by ROC and decision–curve analyses, and prognostic value for progression‐free survival (PFS) was evaluated using Kaplan–Meier analysis. Results: The macro habitat‐based fusion model demonstrated superior performance compared with intratumoral, peritumoral, or single‐habitat models, achieving AUCs of 0.807–0.830 in the external validation cohort. The derived risk score showed a significant association with progression‐free survival (PFS) ( p = 0.011 in the training and p = 0.030 in the validation cohorts). Conclusions: The macro habitat‐based MRI radiomics and deep learning fusion model provides a noninvasive, interpretable, and robust biomarker for predicting treatment response and prognosis in LARC. It holds potential to guide personalized therapeutic strategies, including organ‐preserving approaches and tailored surveillance after NACRT.