Joris M. van Sabben

Joris M. van Sabben

Auto-generated profile, based on publicly available information

Joris M. van Sabben

The Netherlands Cancer Institute 11 Department of Medical Oncology Amsterdam, the Netherlands

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

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

Representative Publications

View All Publications
25 June 2026 in Cancer Research
Verified
Journal Article
Open access
ABArianna Bonetti
Arianna Bonetti
VLVan-Linh Le
Van-Linh Le
ZCZunamys I. Carrero
Zunamys I. Carrero
FWFabian Wolf
Fabian Wolf
MGMarco Gustav
Marco Gustav
JKJakob Nikolas Kather
Jakob Nikolas Kather
Gastrointestinal stromal tumor (GIST) is the most common gastrointestinal mesenchymal tumor, driven by tyrosine–protein kinase (KIT) and platelet-derived growth factor receptor A (PDGFRA) mutations. Specific variants, such as KIT exon 11 deletions, carry prognostic and therapeutic implications, whereas wild-type variants derive limited benefit from tyrosine kinase inhibitors. Given the limited reproducibility of established clinicopathologic risk models, deep learning (DL) applied to whole-slide images (WSI) emerged as a promising tool for molecular classification and prognostic assessment. We analyzed 8398 GIST cases from 21 centers in seven countries, including 7,238 with molecular data and 2,638 with clinical follow-up. DL models were trained on WSIs to predict mutations, treatment sensitivity, and recurrence-free survival (RFS). DL predicted mutational status in GIST from WSIs, with area under the curve of 0.87 for KIT and 0.96 for PDGFRA, and high performance was observed for subtypes, including KIT exon 11 del–inss 557 to 558 (0.67) and PDGFRA exon 18 D842V (0.93). For therapeutic categories, performance reached 0.84 for avapritinib sensitivity and 0.81 for imatinib sensitivity. DL models predicted RFS, with hazard ratios of 8.44 in the overall cohort and 4.74 in patients receiving adjuvant therapy. Prognostic performance was comparable with pathology-based scores, with highest discrimination in the overall cohort and in patients without adjuvant therapy. DL applied to WSIs enables prediction of molecular alterations, treatment sensitivity, and RFS in GIST, performing comparably with established risk scores across international cohorts, providing a baseline for future multimodal predictors. Significance: Deep learning on histology predicts KIT and PDGFRA mutations and stratifies recurrence-free survival in a large international cohort of gastrointestinal stromal tumors from multiple centers.

14 August 2026
Verified
Other
Open access
ABArianna Bonetti
Arianna Bonetti
VLVan-Linh Le
Van-Linh Le
ZCZunamys I. Carrero
Zunamys I. Carrero
FWFabian Wolf
Fabian Wolf
MGMarco Gustav
Marco Gustav
JKJakob Nikolas Kather
Jakob Nikolas Kather
Predictive performance of the DL models for selected molecular mutations in GIST.

14 August 2026
Verified
Other
Open access
ABArianna Bonetti
Arianna Bonetti
VLVan-Linh Le
Van-Linh Le
ZCZunamys I. Carrero
Zunamys I. Carrero
FWFabian Wolf
Fabian Wolf
MGMarco Gustav
Marco Gustav
JKJakob Nikolas Kather
Jakob Nikolas Kather
Statistical pipeline for recurrence-free survival (RFS) analysis. (A) Data partitioning. Boxes with light dashed contours corresponded to the patients with available DL Scores while boxes with solid thick contours to the patient included in the survival analysis. (B) Methodological approach performed in C1, C2 and C3 to obtain deep learning (DL) scores and to develop survival models in the Training cohorts using the Cox proportional hazard (CPH) algorithm and to compare them. Other abbreviations: ext.val: external validation, int. val: internal validation, M.J.: Miettinen-Joensuu. *Covariables were: age, sex, adjuvant TKI therapy and mutational status for C1; and age, sex and mutational status for C2 and C3.

14 August 2026
Verified
Other
Open access
ABArianna Bonetti
Arianna Bonetti
VLVan-Linh Le
Van-Linh Le
ZCZunamys I. Carrero
Zunamys I. Carrero
FWFabian Wolf
Fabian Wolf
MGMarco Gustav
Marco Gustav
JKJakob Nikolas Kather
Jakob Nikolas Kather
Sensitivity, specificity, and F1-score for molecular mutation prediction.

14 August 2026
Verified
Other
Open access
ABArianna Bonetti
Arianna Bonetti
VLVan-Linh Le
Van-Linh Le
ZCZunamys I. Carrero
Zunamys I. Carrero
FWFabian Wolf
Fabian Wolf
MGMarco Gustav
Marco Gustav
JKJakob Nikolas Kather
Jakob Nikolas Kather
Gastrointestinal stromal tumor (GIST) is the most common gastrointestinal mesenchymal tumor, driven by tyrosine–protein kinase (KIT) and platelet-derived growth factor receptor A (PDGFRA) mutations. Specific variants, such as KIT exon 11 deletions, carry prognostic and therapeutic implications, whereas wild-type variants derive limited benefit from tyrosine kinase inhibitors. Given the limited reproducibility of established clinicopathologic risk models, deep learning (DL) applied to whole-slide images (WSI) emerged as a promising tool for molecular classification and prognostic assessment. We analyzed 8398 GIST cases from 21 centers in seven countries, including 7,238 with molecular data and 2,638 with clinical follow-up. DL models were trained on WSIs to predict mutations, treatment sensitivity, and recurrence-free survival (RFS). DL predicted mutational status in GIST from WSIs, with area under the curve of 0.87 for KIT and 0.96 for PDGFRA, and high performance was observed for subtypes, including KIT exon 11 del–inss 557 to 558 (0.67) and PDGFRA exon 18 D842V (0.93). For therapeutic categories, performance reached 0.84 for avapritinib sensitivity and 0.81 for imatinib sensitivity. DL models predicted RFS, with hazard ratios of 8.44 in the overall cohort and 4.74 in patients receiving adjuvant therapy. Prognostic performance was comparable with pathology-based scores, with highest discrimination in the overall cohort and in patients without adjuvant therapy. DL applied to WSIs enables prediction of molecular alterations, treatment sensitivity, and RFS in GIST, performing comparably with established risk scores across international cohorts, providing a baseline for future multimodal predictors. Significance:Deep learning on histology predicts KIT and PDGFRA mutations and stratifies recurrence-free survival in a large international cohort of gastrointestinal stromal tumors from multiple centers.

1 July 2025 in Oncology and Therapy
Verified
Journal Article
Open access
JSJoris M. van Sabben
Joris M. van Sabben
MKMaud B. A. van der Kleij
Maud B. A. van der Kleij
EREvelyne Roets
Evelyne Roets
RTRenaud L. M. Tissier
Renaud L. M. Tissier
DBDorieke E. M. van Balen
Dorieke E. M. van Balen
NSNeeltje Steeghs
Neeltje Steeghs
Following patent expiration of branded imatinib (Glivec®), all Dutch patients with gastrointestinal stromal tumours (GIST) switched from Glivec to generic forms. Following this switch, many patients reported new symptoms. Therefore, we conducted this observational study to assess safety of generic imatinib among patients with GIST in the Netherlands. We included patients with GIST from four hospitals that switched from Glivec to generic imatinib. Within these patients, adverse events (AEs) without the switch to a generic were compared with AEs after the switch using a self-controlled case series design. The reference group was formed by the subset of patients who used imatinib for at least 1 year prior to the switch. As potential causes of increased AEs, we reviewed excipients and analysed plasma trough levels from 1 year prior to 1 year after the switch. In total, 201 patients switched to three generics: Accord® (n = 107), Amarox® (n = 81), and Sandoz® (n = 13). In the reference group (n = 150), 21.3% experienced new AEs, compared with 29.9-34.6% in the different generic groups. All patients that switched to Amarox (odds ratio 2.3; 95% confidence interval (CI): 1.2-4.5) and females that switched to Accord (odds ratio 2.7; 95% CI: 1.1-7.0) experienced a significant increase in AEs. Plasma trough levels were similar among all different formulations. Apart from titanium dioxide in Amarox, no additional excipients were used in any generic form. The transition to generic imatinib in Dutch patients with GIST was safe. After switching to generic imatinib, up to 34.6% of patients experienced new AEs, compared with 21.3% in the reference group, indicating that many AEs may have been mistakenly attributed to the switch. The small increase in AEs that we found was unlikely due to pharmacokinetics or excipients. Therefore, we argue that the nocebo effect, where negative expectations about treatment lead to worsened symptoms, played a large role.