Bastian Grothey
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
Bastian Grothey has a total of 25 co-authors. Here are the top co-authors by number of shared publications:
- Reinhard Büttner
Center for Integrated Oncology Aachen Köln Bonn Düsseldorf, Germany
3 shared publications Thomas Zander
Department of Internal Medicine I University Hospital Cologne, Medical Faculty, University of Cologne Cologne Germany
3 shared publicationsAlexander Quaas
Institute of Pathology, Faculty of Medicine and University Hospital of Cologne, University of Cologne Cologne Germany
2 shared publications- Alexander Quaas
Faculty of Medicine and University Hospital of Cologne, Institute of Pathology, University of Cologne, 50937 Cologne, Germany
2 shared publications Christiane J. Bruns
Department of General, Visceral and Cancer Surgery, Faculty of Medicine University Hospital of Cologne Cologne Germany
2 shared publicationsReinhard Büttner
Institute of Pathology, Faculty of Medicine and University Hospital of Cologne, University of Cologne Cologne Germany
2 shared publications
Latest Publications
- Open Access
- 12 August 2026
- Published by Wiley in International Journal of Cancer
- Vol. 159 (10) , 2595-2605
Sex‐Specific Proteomic Profiling Identifies Pregnancy Zone Protein as a Complement‐Linked Marker of Adverse Prognosis in Esophageal Adenocarcinoma
Esophageal adenocarcinoma (EAC) exhibits marked male predominance with male‐to‐female ratios reaching 8.5:1, yet the molecular basis underlying this sex disparity remains poorly characterized. We analyzed 92 EAC specimens using mass spectrometry–based proteomics, comprising 47 female and 45 male tumors from treatment‐naïve patients. Differential expression, pathway enrichment, immunohistochemical, and survival analyses were used to identify sex‐associated proteomic features and prognostic signatures. Proteomic profiling revealed focal yet biologically meaningful sex differences in EAC. After multiple testing correction, three proteins were differentially expressed: the autosomal protein pregnancy zone protein (PZP), enriched in female tumors, and the Y‐chromosome–encoded proteins DDX3Y and RPS4Y1, which were overexpressed in male tumors. At nominal significance, proteins upregulated in female tumors showed marked enrichment of immune‐related pathways. Proteins correlated with PZP expression formed a network dominated by complement cascade. Clinically, elevated PZP expression was associated with significantly poorer overall survival specifically in male patients, with no significant association detected in the combined or female‐only cohort. Proteome‐wide survival analyses further demonstrated distinct sex‐specific prognostic landscapes. Our study provides the first comprehensive characterization of sex‐associated proteomic differences in EAC. Sex shapes immune‐related pathways, prognostic proteomic signatures, and survival associations. PZP emerges as a sex‐differential, complement‐associated protein with adverse prognostic significance in male patients, highlighting sex as a biologically and clinically relevant variable in EAC.
- 12 August 2026
- Published by Wiley in International Journal of Cancer
- Vol. 159 (10) , 2595-2605
Open AccessSex‐Specific Proteomic Profiling Identifies Pregnancy Zone Protein as a Complement‐Linked Marker of Adverse Prognosis in Esophageal Adenocarcinoma
Esophageal adenocarcinoma (EAC) exhibits marked male predominance with male‐to‐female ratios reaching 8.5:1, yet the molecular basis underlying this sex disparity remains poorly characterized. We analyzed 92 EAC specimens using mass spectrometry–based proteomics, comprising 47 female and 45 male tumors from treatment‐naïve patients. Differential expression, pathway enrichment, immunohistochemical, and survival analyses were used to identify sex‐associated proteomic features and prognostic signatures. Proteomic profiling revealed focal yet biologically meaningful sex differences in EAC. After multiple testing correction, three proteins were differentially expressed: the autosomal protein pregnancy zone protein (PZP), enriched in female tumors, and the Y‐chromosome–encoded proteins DDX3Y and RPS4Y1, which were overexpressed in male tumors. At nominal significance, proteins upregulated in female tumors showed marked enrichment of immune‐related pathways. Proteins correlated with PZP expression formed a network dominated by complement cascade. Clinically, elevated PZP expression was associated with significantly poorer overall survival specifically in male patients, with no significant association detected in the combined or female‐only cohort. Proteome‐wide survival analyses further demonstrated distinct sex‐specific prognostic landscapes. Our study provides the first comprehensive characterization of sex‐associated proteomic differences in EAC. Sex shapes immune‐related pathways, prognostic proteomic signatures, and survival associations. PZP emerges as a sex‐differential, complement‐associated protein with adverse prognostic significance in male patients, highlighting sex as a biologically and clinically relevant variable in EAC.
- Open Access
- 8 July 2026
- Published by Springer Nature in British Journal of Cancer
- Vol. 135 (7) , 1119-1130
Proteomic characterization of SMARCA-deficient esophageal adenocarcinoma reveals complement evasion and DNA repair-associated prognostic subgroups
Background: Oesophageal adenocarcinoma (EAC) is an aggressive malignancy in which the SWI/SNF catalytic subunits SMARCA2 and SMARCA4 are frequently lost, yet the biological and clinical impact of these deficiencies remains unclear. Methods: We analysed 113 EAC specimens using mass spectrometry–based proteomics, including 89 SMARCA-deficient and 24 SMARCA-proficient tumours. Additional immunohistochemical studies were conducted on 98 cases. Differential expression, pathway enrichment, and survival analyses were used to characterise proteomic alterations and identify prognostic features. Results: SMARCA-deficient tumours showed frequent expression of complement-inactivating proteins CD55 and CD59 with distinctive apical and membranous localisation. SMARCA4-deficient patients had significantly poorer survival than both SMARCA2-deficient and SMARCA-proficient patients. Among neoadjuvantly treated SMARCA-deficient tumours, coordinated upregulation of DNA repair proteins separated patients into two prognostic subgroups. The DNA repair–low subgroup had markedly worse survival than both the DNA repair–high subgroup and SMARCA-proficient controls, whereas the DNA repair–high subgroup showed a trend toward improved outcomes. This DNA repair–associated pattern was not observed in neoadjuvantly treated SMARCA-proficient tumours. Conclusions: SMARCA-deficient EAC comprises biologically distinct subgroups with prognostic proteomic signatures. CD55 and CD59 expression may contribute to resistance to antibody-based immunotherapies by complement evasion, while DNA repair profiles could guide post-surgical treatment in neoadjuvantly treated patients.
- 8 July 2026
- Published by Springer Nature in British Journal of Cancer
- Vol. 135 (7) , 1119-1130
Open AccessProteomic characterization of SMARCA-deficient esophageal adenocarcinoma reveals complement evasion and DNA repair-associated prognostic subgroups
Background: Oesophageal adenocarcinoma (EAC) is an aggressive malignancy in which the SWI/SNF catalytic subunits SMARCA2 and SMARCA4 are frequently lost, yet the biological and clinical impact of these deficiencies remains unclear. Methods: We analysed 113 EAC specimens using mass spectrometry–based proteomics, including 89 SMARCA-deficient and 24 SMARCA-proficient tumours. Additional immunohistochemical studies were conducted on 98 cases. Differential expression, pathway enrichment, and survival analyses were used to characterise proteomic alterations and identify prognostic features. Results: SMARCA-deficient tumours showed frequent expression of complement-inactivating proteins CD55 and CD59 with distinctive apical and membranous localisation. SMARCA4-deficient patients had significantly poorer survival than both SMARCA2-deficient and SMARCA-proficient patients. Among neoadjuvantly treated SMARCA-deficient tumours, coordinated upregulation of DNA repair proteins separated patients into two prognostic subgroups. The DNA repair–low subgroup had markedly worse survival than both the DNA repair–high subgroup and SMARCA-proficient controls, whereas the DNA repair–high subgroup showed a trend toward improved outcomes. This DNA repair–associated pattern was not observed in neoadjuvantly treated SMARCA-proficient tumours. Conclusions: SMARCA-deficient EAC comprises biologically distinct subgroups with prognostic proteomic signatures. CD55 and CD59 expression may contribute to resistance to antibody-based immunotherapies by complement evasion, while DNA repair profiles could guide post-surgical treatment in neoadjuvantly treated patients.
- Open Access
- 5 March 2026
- Published by Wiley in International Journal of Cancer
- Vol. 159 (2) , 501-515
Adherens junction protein expression is associated with poor response to neoadjuvant FLOT chemotherapy and pro‐inflammatory tumor microenvironment in esophageal adenocarcinoma
Esophageal adenocarcinoma (EAC) demonstrates poor survival rates despite multimodal treatments, with significant inter‐patient variability in response to standard pre‐/perioperative therapies. This study investigates the relationship between adherens junction (AJ) protein expression in pre‐treatment biopsies and response to neoadjuvant therapies in EAC. Mass spectrometry‐based proteomics was performed on diagnostic biopsies from 157 EAC patients who subsequently received either perioperative FLOT chemotherapy or neoadjuvant CROSS radiochemotherapy. Protein expression profiles were analyzed with emphasis on AJ components. Findings were correlated with treatment response and histopathological parameters. We identified a distinct protein cluster enriched for AJ components that significantly correlated with response to FLOT chemotherapy. Among FLOT major responders, 94% exhibited high AJ expression. Low AJ expression occurred across histological subtypes, with varying frequencies: 27% in tubular‐intestinal, 40% in mixed‐type, and 50% in diffuse carcinomas. Low AJ expression was significantly associated with a pro‐inflammatory tumor microenvironment and extracellular matrix (ECM) remodeling. AJ protein expression represents a potential predictive parameter for FLOT chemotherapy response in EAC and defines distinct phenotypes. The association between reduced AJ expression and inflammatory features suggests these tumors may represent candidates for further investigation of targeted therapeutic approaches, though the role of immunotherapy in this context remains to be determined. Pre‐treatment AJ assessment could guide therapy selection to avoid ineffective therapies.
- 5 March 2026
- Published by Wiley in International Journal of Cancer
- Vol. 159 (2) , 501-515
Open AccessAdherens junction protein expression is associated with poor response to neoadjuvant FLOT chemotherapy and pro‐inflammatory tumor microenvironment in esophageal adenocarcinoma
Esophageal adenocarcinoma (EAC) demonstrates poor survival rates despite multimodal treatments, with significant inter‐patient variability in response to standard pre‐/perioperative therapies. This study investigates the relationship between adherens junction (AJ) protein expression in pre‐treatment biopsies and response to neoadjuvant therapies in EAC. Mass spectrometry‐based proteomics was performed on diagnostic biopsies from 157 EAC patients who subsequently received either perioperative FLOT chemotherapy or neoadjuvant CROSS radiochemotherapy. Protein expression profiles were analyzed with emphasis on AJ components. Findings were correlated with treatment response and histopathological parameters. We identified a distinct protein cluster enriched for AJ components that significantly correlated with response to FLOT chemotherapy. Among FLOT major responders, 94% exhibited high AJ expression. Low AJ expression occurred across histological subtypes, with varying frequencies: 27% in tubular‐intestinal, 40% in mixed‐type, and 50% in diffuse carcinomas. Low AJ expression was significantly associated with a pro‐inflammatory tumor microenvironment and extracellular matrix (ECM) remodeling. AJ protein expression represents a potential predictive parameter for FLOT chemotherapy response in EAC and defines distinct phenotypes. The association between reduced AJ expression and inflammatory features suggests these tumors may represent candidates for further investigation of targeted therapeutic approaches, though the role of immunotherapy in this context remains to be determined. Pre‐treatment AJ assessment could guide therapy selection to avoid ineffective therapies.
- Top Cited PapersOpen Access
- 31 March 2025
- Published by Springer Nature in Communications Medicine
- Vol. 5 (1) , 1-13
Comprehensive testing of large language models for extraction of structured data in pathology
Background: Pathology departments generate large volumes of unstructured data as free-text diagnostic reports. Converting these reports into structured formats for analytics or artificial intelligence projects requires substantial manual effort by specialized personnel. While recent studies show promise in using advanced language models for structuring pathology data, they primarily rely on proprietary models, raising cost and privacy concerns. Additionally, important aspects such as prompt engineering and model quantization for deployment on consumer-grade hardware remain unaddressed. Methods: We created a dataset of 579 annotated pathology reports in German and English versions. Six language models (proprietary: GPT-4; open-source: Llama2 13B, Llama2 70B, Llama3 8B, Llama3 70B, and Qwen2.5 7B) were evaluated for their ability to extract eleven key parameters from these reports. Additionally, we investigated model performance across different prompt engineering strategies and model quantization techniques to assess practical deployment scenarios. Results: Here we show that open-source language models extract structured data from pathology reports with high precision, matching the accuracy of proprietary GPT-4 model. The precision varies significantly across different models and configurations. These variations depend on specific prompt engineering strategies and quantization methods used during model deployment. Conclusions: Open-source language models demonstrate comparable performance to proprietary solutions in structuring pathology report data. This finding has significant implications for healthcare institutions seeking cost-effective, privacy-preserving data structuring solutions. The variations in model performance across different configurations provide valuable insights for practical deployment in pathology departments. Our publicly available bilingual dataset serves as both a benchmark and a resource for future research.
- 31 March 2025
- Published by Springer Nature in Communications Medicine
- Vol. 5 (1) , 1-13
Top Cited PapersOpen AccessComprehensive testing of large language models for extraction of structured data in pathology
Background: Pathology departments generate large volumes of unstructured data as free-text diagnostic reports. Converting these reports into structured formats for analytics or artificial intelligence projects requires substantial manual effort by specialized personnel. While recent studies show promise in using advanced language models for structuring pathology data, they primarily rely on proprietary models, raising cost and privacy concerns. Additionally, important aspects such as prompt engineering and model quantization for deployment on consumer-grade hardware remain unaddressed. Methods: We created a dataset of 579 annotated pathology reports in German and English versions. Six language models (proprietary: GPT-4; open-source: Llama2 13B, Llama2 70B, Llama3 8B, Llama3 70B, and Qwen2.5 7B) were evaluated for their ability to extract eleven key parameters from these reports. Additionally, we investigated model performance across different prompt engineering strategies and model quantization techniques to assess practical deployment scenarios. Results: Here we show that open-source language models extract structured data from pathology reports with high precision, matching the accuracy of proprietary GPT-4 model. The precision varies significantly across different models and configurations. These variations depend on specific prompt engineering strategies and quantization methods used during model deployment. Conclusions: Open-source language models demonstrate comparable performance to proprietary solutions in structuring pathology report data. This finding has significant implications for healthcare institutions seeking cost-effective, privacy-preserving data structuring solutions. The variations in model performance across different configurations provide valuable insights for practical deployment in pathology departments. Our publicly available bilingual dataset serves as both a benchmark and a resource for future research.
- Open Access
- 19 January 2022
- Published by Springer Nature in npj Systems Biology and Applications
- Vol. 8 (1) , 1-11
Diffusion kernel-based predictive modeling of KRAS dependency in KRAS wild type cancer cell lines
Recent progress in clinical development of KRAS inhibitors has raised interest in predicting the tumor dependency on frequently mutated RAS-pathway oncogenes. However, even without such activating mutations, RAS proteins represent core components in signal integration of several membrane-bound kinases. This raises the question of applications of specific inhibitors independent from the mutational status. Here, we examined CRISPR/RNAi data from over 700 cancer cell lines and identified a subset of cell lines without KRAS gain-of-function mutations (KRASwt) which are dependent on KRAS expression. Combining machine learning-based modeling and whole transcriptome data with prior variable selection through protein-protein interaction network analysis by a diffusion kernel successfully predicted KRAS dependency in the KRASwtsubgroup and in all investigated cancer cell lines. In contrast, modeling by RAS activating events (RAE) or previously published RAS RNA-signatures did not provide reliable results, highlighting the heterogeneous distribution of RAE in KRASwtcell lines and the importance of methodological references for expression signature modeling. Furthermore, we show that predictors of KRASwtmodels contain non-substitutable information signals, indicating a KRAS dependency phenotype in the KRASwtsubgroup. Our data suggest that KRAS dependent cancers harboring KRAS wild type status could be targeted by directed therapeutic approaches. RNA-based machine learning models could help in identifying responsive and non-responsive tumors.
- 19 January 2022
- Published by Springer Nature in npj Systems Biology and Applications
- Vol. 8 (1) , 1-11
Open AccessDiffusion kernel-based predictive modeling of KRAS dependency in KRAS wild type cancer cell lines
Recent progress in clinical development of KRAS inhibitors has raised interest in predicting the tumor dependency on frequently mutated RAS-pathway oncogenes. However, even without such activating mutations, RAS proteins represent core components in signal integration of several membrane-bound kinases. This raises the question of applications of specific inhibitors independent from the mutational status. Here, we examined CRISPR/RNAi data from over 700 cancer cell lines and identified a subset of cell lines without KRAS gain-of-function mutations (KRASwt) which are dependent on KRAS expression. Combining machine learning-based modeling and whole transcriptome data with prior variable selection through protein-protein interaction network analysis by a diffusion kernel successfully predicted KRAS dependency in the KRASwtsubgroup and in all investigated cancer cell lines. In contrast, modeling by RAS activating events (RAE) or previously published RAS RNA-signatures did not provide reliable results, highlighting the heterogeneous distribution of RAE in KRASwtcell lines and the importance of methodological references for expression signature modeling. Furthermore, we show that predictors of KRASwtmodels contain non-substitutable information signals, indicating a KRAS dependency phenotype in the KRASwtsubgroup. Our data suggest that KRAS dependent cancers harboring KRAS wild type status could be targeted by directed therapeutic approaches. RNA-based machine learning models could help in identifying responsive and non-responsive tumors.