ESMO Real World Data and Digital Oncology最新文献

{"title":"Portuguese National Cancer Registry: from local registries to high-quality nationwide real-world data","authors":"T.M. Garcia, A. Sousa, P. Leite-Silva, L. Lopes-Conceição, O. Godinho, M.J. Bento","doi":"10.1016/j.esmorw.2026.100757","DOIUrl":"10.1016/j.esmorw.2026.100757","url":null,"abstract":"<div><ul><li><span>•</span><span><div>RON introduced new clinical variables, refining tumor and treatment details.</div></span></li><li><span>•</span><span><div>Quality challenges persist owing to manual data entry.</div></span></li><li><span>•</span><span><div>Future investments include automated integration, validation, phased data release.</div></span></li></ul></div><div><h3>Background</h3><div>Portugal’s population-based cancer registries have evolved from fragmented municipal initiatives, starting in 1976, to full national coverage by 2007, culminating in the legislative creation of the National Cancer Registry [Registo Oncológico Nacional (RON)] in 2017.</div></div><div>Portugal’s population-based cancer registries have evolved from fragmented municipal initiatives, starting in 1976, to full national coverage by 2007, culminating in the legislative creation of the National Cancer Registry [Registo Oncológico Nacional (RON)] in 2017.</div><div><h3>Patients and methods</h3><div>This short communication describes RON’s structural evolution, data quality, research utility, and readiness for international reporting.</div></div><div>This short communication describes RON’s structural evolution, data quality, research utility, and readiness for international reporting.</div><div><h3>Results</h3><div>Although regional registries captured 20 core epidemiological variables, RON introduced >30 new clinical variables aligned with European Network of Cancer Registries recommendations, refining tumor characterization and treatment details. Analysis of 2022 data reveals that core incidence variables are 100% complete, with microscopic verification (97.8%), death certificate only (<0.1%), and unknown basis of diagnosis (<0.1%) meeting international benchmarks. However, challenges persist: stage is missing in 40.4% of cases, ill-defined morphology or topography affects 3.5%, unrelated surgical codes affect 14.9%, and poor stage standardization is evidenced by 779 recorded options. Despite partial automations, manual data entry remains substantial, contributing to a 3-year publication lag. Nevertheless, RON has supported 56 national and international studies since 2018, with data now sufficient for consideration in international reports.</div></div><div>Although regional registries captured 20 core epidemiological variables, RON introduced >30 new clinical variables aligned with European Network of Cancer Registries recommendations, refining tumor characterization and treatment details. Analysis of 2022 data reveals that core incidence variables are 100% complete, with microscopic verification (97.8%), death certificate only (<0.1%), and unknown basis of diagnosis (<0.1%) meeting international benchmarks. However, challenges persist: stage is missing in 40.4% of cases, ill-defined morphology or topography affects 3.5%, unrelated surgical codes affect 14.9%, and poor stage standardization is evidenced by 779 recorded options. Despite partia","PeriodicalId":100491,"journal":{"name":"ESMO Real World Data and Digital Oncology","volume":"13 ","pages":"Article 100757"},"PeriodicalIF":2.6,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148951079","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Safety, treatment delivery, and pathological response before and after pembrolizumab implementation for early ER-negative, HER2-normal breast cancer","authors":"A.F. Henriksen, A.N.H. Lundgren, M.A. Andersen, T.A. Gerds, A.S. Knop, T.S. Petersen, U. Lassen, A. Bjerrum","doi":"10.1016/j.esmorw.2026.100769","DOIUrl":"10.1016/j.esmorw.2026.100769","url":null,"abstract":"<div><ul><li><span>•</span><span><div>EHRs enabled detailed NST assessment in early ER-negative, HER2-normal breast cancer.</div></span></li><li><span>•</span><span><div>pCR increased from 49% to 68% after pembrolizumab entered routine NST.</div></span></li><li><span>•</span><span><div>Grade 3/4 toxicity was more frequent after implementation (71% versus 65%).</div></span></li><li><span>•</span><span><div>Suspected severe irAEs occurred in 21% of pembrolizumab-treated patients.</div></span></li><li><span>•</span><span><div>Complete AE ascertainment required manual review of unstructured EHR data.</div></span></li></ul></div><div><h3>Background</h3><div>Pembrolizumab has become standard neoadjuvant systemic therapy (NST) for early estrogen receptor (ER)-negative, human epidermal growth factor receptor 2 (HER2)-normal early breast cancer (eBC). Electronic health record (EHR) data may support real-world assessment of its implementation in routine care.</div></div><div>Pembrolizumab has become standard neoadjuvant systemic therapy (NST) for early estrogen receptor (ER)-negative, human epidermal growth factor receptor 2 (HER2)-normal early breast cancer (eBC). Electronic health record (EHR) data may support real-world assessment of its implementation in routine care.</div><div><h3>Patients and methods</h3><div>Using centralized EHR data from eastern Denmark, we identified adults with ER-negative, HER2-normal eBC receiving NST. Patients were grouped by treatment start before (April 2021-June 2022; comparator) or after (April 2023-June 2024; study) pembrolizumab implementation. We compared treatment exposure, adverse events (AEs) from laboratory data, clinical notes, and AE-indicating medications, and pathological complete response (pCR) from pathology records. Follow-up ended 200 days after surgery, progression, or death. Manual validation included 50 patients.</div></div><div>Using centralized EHR data from eastern Denmark, we identified adults with ER-negative, HER2-normal eBC receiving NST. Patients were grouped by treatment start before (April 2021-June 2022; comparator) or after (April 2023-June 2024; study) pembrolizumab implementation. We compared treatment exposure, adverse events (AEs) from laboratory data, clinical notes, and AE-indicating medications, and pathological complete response (pCR) from pathology records. Follow-up ended 200 days after surgery, progression, or death. Manual validation included 50 patients.</div><div><h3>Results</h3><div>Among 210 included patients, 109 were in the study population and 101 in the comparator population. Baseline characteristics were broadly similar. Pembrolizumab was received by 77% versus 0% and carboplatin omission was less frequent after implementation (7% versus 47%). During NST, grade 3/4 AEs were observed in 77 study patients (71%) and 66 comparator patients (65%). Suspected grade 3/4 autoimmune disorders during NST occurred in 18 pembrolizumab-treated study patients (21%) and in no comparator ","PeriodicalId":100491,"journal":{"name":"ESMO Real World Data and Digital Oncology","volume":"13 ","pages":"Article 100769"},"PeriodicalIF":2.6,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148951077","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Real-world data sources for pediatric cancers: a targeted literature review","authors":"P.J. Kulkarni, T.T. Vo, X. Liu, S.J. Diede, M. Burcu","doi":"10.1016/j.esmorw.2026.100766","DOIUrl":"10.1016/j.esmorw.2026.100766","url":null,"abstract":"<div><ul><li><span>•</span><span><div>RWD can advance precision medicine and contextualize trial result.</div></span></li><li><span>•</span><span><div>RWD sources are mainly population-based, although data element completeness varies.</div></span></li><li><span>•</span><span><div>Integration of genomic data into RWD would add value to pediatric cancer research.</div></span></li></ul></div><div><h3>Background</h3><div>Pediatric cancers cause substantial childhood mortality and evidence gaps remain for rare tumors. Real-world data (RWD) can complement trials by providing long-term follow-up, broader populations, and external controls. We reviewed global pediatric cancer RWD to characterize sources and key data elements.</div></div><div>Pediatric cancers cause substantial childhood mortality and evidence gaps remain for rare tumors. Real-world data (RWD) can complement trials by providing long-term follow-up, broader populations, and external controls. We reviewed global pediatric cancer RWD to characterize sources and key data elements.</div><div><h3>Materials and methods</h3><div>PubMed was searched (February 2015-February 2025) for studies referencing cancer databases or initiatives collecting RWD on patients aged 0-25 years. After screening 1666 records, 225 articles yielded 134 unique data sources. We extracted database type, geography, population, tumor/treatment coverage, and availability of demographic, clinical, and outcome variables.</div></div><div>PubMed was searched (February 2015-February 2025) for studies referencing cancer databases or initiatives collecting RWD on patients aged 0-25 years. After screening 1666 records, 225 articles yielded 134 unique data sources. We extracted database type, geography, population, tumor/treatment coverage, and availability of demographic, clinical, and outcome variables.</div><div><h3>Results</h3><div>Of 134 sources, 34.1% were pediatric-only. Most were in Europe (47.0%) or the Americas (18.7%). About 75% were population-based registries and ∼10% genomic/biobank resources. Age and sex were documented in >60% of sources; tumor site, histology, and stage in >50%; tumor grade in ∼30%; treatment modalities in 36%-40%; and survival in 61.2%. Race, medical history, and clinical outcomes (progression and response) were infrequently reported.</div></div><div>Of 134 sources, 34.1% were pediatric-only. Most were in Europe (47.0%) or the Americas (18.7%). About 75% were population-based registries and ∼10% genomic/biobank resources. Age and sex were documented in >60% of sources; tumor site, histology, and stage in >50%; tumor grade in ∼30%; treatment modalities in 36%-40%; and survival in 61.2%. Race, medical history, and clinical outcomes (progression and response) were infrequently reported.</div><div><h3>Conclusions</h3><div>Pediatric oncology RWD are numerous but concentrated in high-income regions and vary in completeness. Core demographics and major tumor features are commonly recorded yet s","PeriodicalId":100491,"journal":{"name":"ESMO Real World Data and Digital Oncology","volume":"13 ","pages":"Article 100766"},"PeriodicalIF":2.6,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148934692","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Real-world characteristics and invasive disease-free survival of resected patients with HR-positive/HER2-negative early breast cancer: evidence from German population-based cancer registries","authors":"P. Grieger, M.Y. Beykloo, C. Lee, J. Moreira, N. Niklas, D. Twardella, B. Köpf, D. Reil, C. Schneider, S. Luttmann, A. Nennecke, S.-Z. Kim-Wanner, F. Oesterling, B. Holleczek, R. Pritzkuleit, A. Katalinic, A. Waldmann","doi":"10.1016/j.esmorw.2026.100765","DOIUrl":"10.1016/j.esmorw.2026.100765","url":null,"abstract":"<div><ul><li><span>•</span><span><div>Among patients with 43 214 HR-positive/HER2-negative eBC, 2142 experienced recurrence during follow-up.</div></span></li><li><span>•</span><span><div>Recurrence was more frequent in male, premenopausal patients with advanced disease.</div></span></li><li><span>•</span><span><div>The 3- and 5-year iDFS were 90.1% and 82.9%.</div></span></li><li><span>•</span><span><div>iDFS varied across subgroups, with lower rates in higher-stage patients.</div></span></li><li><span>•</span><span><div>Key prognostic factors for iDFS were confirmed in a German real-world cohort.</div></span></li></ul></div><div><h3>Background</h3><div>Hormone receptor-positive (HR-positive) and human epidermal growth factor receptor 2-negative (HER2-negative) tumors are the most prevalent subtypes of early breast cancer (eBC). Despite the generally favorable prognosis, a relevant proportion experiences invasive disease events following surgery. We aimed to describe characteristics, treatment modalities, and survival outcomes of surgically treated patients with HR-positive/HER2-negative eBC in Germany.</div></div><div>Hormone receptor-positive (HR-positive) and human epidermal growth factor receptor 2-negative (HER2-negative) tumors are the most prevalent subtypes of early breast cancer (eBC). Despite the generally favorable prognosis, a relevant proportion experiences invasive disease events following surgery. We aimed to describe characteristics, treatment modalities, and survival outcomes of surgically treated patients with HR-positive/HER2-negative eBC in Germany.</div><div><h3>Materials and methods</h3><div>This study used data from eight population-based clinical cancer registries in Germany. Patients diagnosed between 2017 and 2021 with HR-positive/HER2-negative eBC who underwent surgery within 4 months of diagnosis were included. Patient and tumor characteristics and treatment received before recurrence were described. Kaplan–Meier statistics and multivariable Cox regression were used to assess invasive disease-free survival (iDFS).</div></div><div>This study used data from eight population-based clinical cancer registries in Germany. Patients diagnosed between 2017 and 2021 with HR-positive/HER2-negative eBC who underwent surgery within 4 months of diagnosis were included. Patient and tumor characteristics and treatment received before recurrence were described. Kaplan–Meier statistics and multivariable Cox regression were used to assess invasive disease-free survival (iDFS).</div><div><h3>Results</h3><div>A total of 43 214 patients with a median follow-up time of 47 months were included. Among these, 5635 experienced an iDFS event, of which 38.0% were recurrences as the first observed event. Recurrence more often occurred among men, premenopausal women, and patients with larger tumors, greater nodal involvement, and higher tumor grade. The 5-year iDFS was 82.9% (95% confidence interval 82.4-83.3). In multivariable analysis, higher age","PeriodicalId":100491,"journal":{"name":"ESMO Real World Data and Digital Oncology","volume":"13 ","pages":"Article 100765"},"PeriodicalIF":2.6,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148951074","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"From rejection to formative development: recalibrating decisions for under-resourced authors","authors":"H.A. Tafenzi, R. Belbaraka","doi":"10.1016/j.esmorw.2026.100768","DOIUrl":"10.1016/j.esmorw.2026.100768","url":null,"abstract":"<div><ul><li><span>•</span><span><div>Mentorship escrow converts borderline submissions into time-limited remediation.</div></span></li><li><span>•</span><span><div>Integrates AI diagnostics, targeted methodological checks, and development programmes.</div></span></li><li><span>•</span><span><div>Produces concise mentor notes to inform adjudication without replacing peer review.</div></span></li><li><span>•</span><span><div>Preserves reviewer effort via secure transfer and structured development packages.</div></span></li><li><span>•</span><span><div>Recommends staged pilots, mentor accreditation, and equity metrics for scale-up.</div></span></li></ul></div><div><div>Peer review and editorial triage together are the cornerstones of scientific credibility yet their combined cracks are increasingly visible. As real-world data and digital health technologies have become increasingly important sources of evidence generation, many promising submissions from under-resourced settings receive terse desk rejections or nonspecialist reviews that obscure remediable methodological and reporting issues, limiting learning and career progression. We propose a compact, author-initiated, equity-centred remediation pathway—the mentorship escrow—that integrates upstream supports, a core remediation window, and a downstream pathway, converting borderline submissions into time-bounded and capacity-building interventions. We outline practical implementation steps, pilot design considerations, and measurable equity and quality metrics; discuss trade-offs across components; and address variable constraints including mentor availability, funding, potential selection bias, and risks of dependency, proposing mitigation strategies to ensure scalable, sustainable capacity building that preserves editorial standards while broadening the circulation of locally relevant evidence.</div></div><div>Peer review and editorial triage together are the cornerstones of scientific credibility yet their combined cracks are increasingly visible. As real-world data and digital health technologies have become increasingly important sources of evidence generation, many promising submissions from under-resourced settings receive terse desk rejections or nonspecialist reviews that obscure remediable methodological and reporting issues, limiting learning and career progression. We propose a compact, author-initiated, equity-centred remediation pathway—the mentorship escrow—that integrates upstream supports, a core remediation window, and a downstream pathway, converting borderline submissions into time-bounded and capacity-building interventions. We outline practical implementation steps, pilot design considerations, and measurable equity and quality metrics; discuss trade-offs across components; and address variable constraints including mentor availability, funding, potential selection bias, and risks of dependency, proposing mitigation strategies to ensure scalable, sustainable capacity build","PeriodicalId":100491,"journal":{"name":"ESMO Real World Data and Digital Oncology","volume":"13 ","pages":"Article 100768"},"PeriodicalIF":2.6,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148951078","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Extended cancer registry data for real-world evidence in bladder cancer: an update on the Prospective Bladder Cancer Infrastructure","authors":"F.M. Uunk, J. Bosveld, J.L. Boormans, A.G. van der Heijden, L.A. Kiemeney, N. Mehra, R.P. Meijer, B.B.M. Suelmann, E.P.A. Janssen, A. Richters, K.K.H. Aben","doi":"10.1016/j.esmorw.2026.100772","DOIUrl":"10.1016/j.esmorw.2026.100772","url":null,"abstract":"<div><ul><li><span>•</span><span><div>ProBCI is a nationwide BCa cohort embedded in the NCR.</div></span></li><li><span>•</span><span><div>The population-based full cohort includes longitudinal data for all patients with BCa.</div></span></li><li><span>•</span><span><div>The informed consent-based active cohort also includes PROMs and biomaterials.</div></span></li><li><span>•</span><span><div>Collected data support diverse research, e.g. epidemiological and biomarker studies.</div></span></li><li><span>•</span><span><div>ProBCI shows the value of integrated registry-based data for BCa research.</div></span></li></ul></div><div><h3>Background</h3><div>To advance clinical research and innovation in bladder cancer care, the Dutch Prospective Bladder Cancer Infrastructure (ProBCI) was established as a nationwide platform and has now been operational for 5 years.</div></div><div>To advance clinical research and innovation in bladder cancer care, the Dutch Prospective Bladder Cancer Infrastructure (ProBCI) was established as a nationwide platform and has now been operational for 5 years.</div><div><h3>Materials and methods</h3><div>ProBCI is a population-based cohort including all patients diagnosed since 2020 with high-risk non-muscle-invasive bladder cancer (HR-NMIBC), muscle-invasive bladder cancer (MIBC), or metastatic bladder cancer (mBC) in the Netherlands, comprising a Full and an Active Cohort. The full cohort includes all eligible patients, from whom core (HR-NMIBC and MIBC) or extended (mBC) clinical data and follow-up are collected within the Netherlands Cancer Registry (NCR). The active cohort is a subset of the full cohort and consists of patients who provided informed consent for additional collection of patient-reported outcomes and biomaterials alongside detailed clinical data and follow-up until death.</div></div><div>ProBCI is a population-based cohort including all patients diagnosed since 2020 with high-risk non-muscle-invasive bladder cancer (HR-NMIBC), muscle-invasive bladder cancer (MIBC), or metastatic bladder cancer (mBC) in the Netherlands, comprising a Full and an Active Cohort. The full cohort includes all eligible patients, from whom core (HR-NMIBC and MIBC) or extended (mBC) clinical data and follow-up are collected within the Netherlands Cancer Registry (NCR). The active cohort is a subset of the full cohort and consists of patients who provided informed consent for additional collection of patient-reported outcomes and biomaterials alongside detailed clinical data and follow-up until death.</div><div><h3>Results</h3><div>As of 1 January 2026, the ProBCI full cohort comprised 20 737 patients. A growing number of participating hospitals (now 24) have been prospectively recruiting patients in the active cohort, with 1519 patients included to date. Among these, 73% completed baseline questionnaires (optional component) and 91% provided blood samples. Data from both cohorts are used by various studies including population ch","PeriodicalId":100491,"journal":{"name":"ESMO Real World Data and Digital Oncology","volume":"13 ","pages":"Article 100772"},"PeriodicalIF":2.6,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148934363","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"X-CART: a multimodal real-world data pipeline for explainable AI-driven patient navigation in radiation oncology","authors":"J.A. Zink, R. Rashid, L. Chinthala, S. Kheirinejad, B.M. White, P. Kheirkhah Rahimabad, S. Hashtarkhani, F.A. Kumsa, C.L. Brett, R.L. Davis, D.L. Schwartz, A. Shaban-Nejad","doi":"10.1016/j.esmorw.2026.100771","DOIUrl":"10.1016/j.esmorw.2026.100771","url":null,"abstract":"<div><ul><li><span>•</span><span><div>Introduces a multimodal RWD pipeline for radiation oncology.</div></span></li><li><span>•</span><span><div>Integrates EHR, geospatial, and SDoH data into an AI-ready framework.</div></span></li><li><span>•</span><span><div>Enables explainable AI for predicting RTIs.</div></span></li><li><span>•</span><span><div>Designed as a registry-extensible architecture supporting interoperability.</div></span></li><li><span>•</span><span><div>Enables registry-based RWE, including comparative effectiveness analyses.</div></span></li></ul></div><div><h3>Background</h3><div>Multimodal explainable artificial intelligence (XAI) is transforming oncology, yet its clinical adoption is constrained by limited interpretability and fragmented use of real-world data (RWD), including electronic medical records, geospatial data, and social determinants of health. We present the XAI CAncer Radiation Therapy (X-CART) platform, a multimodal RWD pipeline designed to enable explainable, clinically actionable precision oncology.</div></div><div>Multimodal explainable artificial intelligence (XAI) is transforming oncology, yet its clinical adoption is constrained by limited interpretability and fragmented use of real-world data (RWD), including electronic medical records, geospatial data, and social determinants of health. We present the XAI CAncer Radiation Therapy (X-CART) platform, a multimodal RWD pipeline designed to enable explainable, clinically actionable precision oncology.</div><div><h3>Materials and methods</h3><div>X-CART integrates heterogeneous RWD into a unified, AI-ready architecture for explainable radiotherapy interruption (RTI) risk prediction and patient navigation. A retrospective proof-of-implementation analysis demonstrated the pipeline’s predictive modeling and explainability. This framework supports targeted, patient-level intervention strategies.</div></div><div>X-CART integrates heterogeneous RWD into a unified, AI-ready architecture for explainable radiotherapy interruption (RTI) risk prediction and patient navigation. A retrospective proof-of-implementation analysis demonstrated the pipeline’s predictive modeling and explainability. This framework supports targeted, patient-level intervention strategies.</div><div><h3>Results</h3><div>We developed a scalable, interoperable pipeline that synthesizes clinical and community-level data into a secure multimodal framework for RTI risk prediction and patient navigation. The system employs iterative machine learning with feedback-driven retraining and delivers real-time, interpretable outputs through an electronic medical record-embedded dashboard. The architecture is registry-extensible and aligned with interoperability standards to support multi-institutional deployment. As proof of implementation of the Data Science module, a retrospective cohort analysis including 2525 patients identified 622 RTI events (24.6%) and demonstrated modest discrimination with eXtreme Gradi","PeriodicalId":100491,"journal":{"name":"ESMO Real World Data and Digital Oncology","volume":"13 ","pages":"Article 100771"},"PeriodicalIF":2.6,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148934364","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Primary care physician documentation in oncology electronic health records: prognostic marker or immortal-time bias?","authors":"P.E. Heudel, J.-Y. Blay","doi":"10.1016/j.esmorw.2026.100773","DOIUrl":"10.1016/j.esmorw.2026.100773","url":null,"abstract":"<div><ul><li><span>•</span><span><div>Documentation showed survival benefit; association lost after adjustment for timing.</div></span></li><li><span>•</span><span><div>Advantage driven by time-related bias; better outcomes with later documentation.</div></span></li><li><span>•</span><span><div>Temporal alignment showed PCP documentation had no independent prognostic value.</div></span></li><li><span>•</span><span><div>Time-dependent EHR variables require proper handling for valid causal inference.</div></span></li></ul></div><div><h3>Background</h3><div>Structured administrative fields in oncology electronic health records (EHRs) are used in studies but their interpretation may depend on when they are completed. We assessed whether dated primary care physician (PCP) documentation was associated with survival after accounting for documentation timing.</div></div><div>Structured administrative fields in oncology electronic health records (EHRs) are used in studies but their interpretation may depend on when they are completed. We assessed whether dated primary care physician (PCP) documentation was associated with survival after accounting for documentation timing.</div><div><h3>Materials and methods</h3><div>We conducted a retrospective cohort study using routinely collected oncology EHR data from a single French comprehensive cancer centre (ConSoRe platform), including patients treated for cancer between 2015 and 2017. Survival was analysed from cancer care initiation to death from any cause or administrative end of follow-up; no patients were censored before 12 months. Conventional analyses compared patients with or without an exploitable dated PCP declaration using inverse probability of treatment weighting (IPTW) targeting the average treatment effect based on age, sex, metastatic status, and tumour location. Temporally aligned analyses included baseline-available exposure definitions, a time-dependent Cox model, and a 6-month landmark analysis.</div></div><div>We conducted a retrospective cohort study using routinely collected oncology EHR data from a single French comprehensive cancer centre (ConSoRe platform), including patients treated for cancer between 2015 and 2017. Survival was analysed from cancer care initiation to death from any cause or administrative end of follow-up; no patients were censored before 12 months. Conventional analyses compared patients with or without an exploitable dated PCP declaration using inverse probability of treatment weighting (IPTW) targeting the average treatment effect based on age, sex, metastatic status, and tumour location. Temporally aligned analyses included baseline-available exposure definitions, a time-dependent Cox model, and a 6-month landmark analysis.</div><div><h3>Results</h3><div>Among the 15 216 patients, 10 418 had a dated PCP declaration (68.5%), 4757 had a declared PCP but no exploitable date (31.2%), and 41 had no PCP declared (0.3%). During the median follow-up of 102.1 months, 7175 ","PeriodicalId":100491,"journal":{"name":"ESMO Real World Data and Digital Oncology","volume":"13 ","pages":"Article 100773"},"PeriodicalIF":2.6,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148951076","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Impact of the COVID-19 pandemic on cancer incidence, stage, mortality, and short-term survival in Ireland: a population-based analysis","authors":"A. Brennan, J. McDevitt, H. Heidarian Miri, S. O’Reilly, L. Burke, K. Bennett, M.T. Redaniel, D. Murray","doi":"10.1016/j.esmorw.2026.100767","DOIUrl":"10.1016/j.esmorw.2026.100767","url":null,"abstract":"<div><ul><li><span>•</span><span><div>No consistent rise in late-stage cancers was observed following the pandemic.</div></span></li><li><span>•</span><span><div>Survival rates for all cancer sites in 2020-2021 remained in line with previous years.</div></span></li><li><span>•</span><span><div>Sharp drop in cancers in early 2020 as COVID-19 disrupted diagnostic services.</div></span></li><li><span>•</span><span><div>Systems adapted rapidly and diagnoses returned to expected levels by 2022.</div></span></li><li><span>•</span><span><div>Real-world evidence on the impact of COVID-19 on cancer control in Ireland.</div></span></li></ul></div><div><h3>Background</h3><div>Concern exists that disruptions to health services in combination with changed health-seeking behaviour as a consequence of the coronavirus disease (COVID-19) pandemic would negatively impact cancer outcomes. The purpose of this study was to assess how the COVID-19 pandemic affected cancer incidence, stage at diagnosis, mortality, and 1-year net survival in Ireland.</div></div><div>Concern exists that disruptions to health services in combination with changed health-seeking behaviour as a consequence of the coronavirus disease (COVID-19) pandemic would negatively impact cancer outcomes. The purpose of this study was to assess how the COVID-19 pandemic affected cancer incidence, stage at diagnosis, mortality, and 1-year net survival in Ireland.</div><div><h3>Materials and methods</h3><div>Age-standardised incidence and mortality rates were generated for three periods (2018-2019, 2020-2021, and 2022). Quarterly age-standardised incidence rates, incidence rate ratios (IRRs), and percentage change (PC) by cancer site were generated for Q1 2020 to Q4 2021. Annual incidence rates, IRRs and PC by stage at diagnosis were estimated. Age- and stage-adjusted 1-year net survival was estimated using the Pohar–Perme method.</div></div><div>Age-standardised incidence and mortality rates were generated for three periods (2018-2019, 2020-2021, and 2022). Quarterly age-standardised incidence rates, incidence rate ratios (IRRs), and percentage change (PC) by cancer site were generated for Q1 2020 to Q4 2021. Annual incidence rates, IRRs and PC by stage at diagnosis were estimated. Age- and stage-adjusted 1-year net survival was estimated using the Pohar–Perme method.</div><div><h3>Results</h3><div>The age-standardised quarterly incidence rate of all invasive cancers (excluding nonmelanoma skin cancer) decreased significantly in Q2 2020, by 27% (95% CI −42% to −7%, <em>P</em> = 0.01), compared with 2018-2019. Decreases in the incidence of early-stage cancers (stages I or II) were seen in 2020, particularly for cancers in which organised or opportunistic screening occurs. There was little evidence of consistent increases in the incidence rate of stage III or IV cancers in 2021 or 2022. Age- and stage-adjusted 1-year net survival for all cancer sites in 2020-2021 remained in line with previous years.</div><","PeriodicalId":100491,"journal":{"name":"ESMO Real World Data and Digital Oncology","volume":"13 ","pages":"Article 100767"},"PeriodicalIF":2.6,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148934812","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Development and validation of a human-supervised AI-augmented living oncology evidence platform: a breast cancer pilot study☆","authors":"H. Rugo, A. Forsythe, D. Flora, S. Glück, S. Grieve, R. Campden, R.J. Liu, M. Musat, J.M. Rege, P.A. Kaufman, R.L. Coleman, P. Tarantino, J. O’Shaughnessy, L. Schwartzberg","doi":"10.1016/j.esmorw.2026.100770","DOIUrl":"10.1016/j.esmorw.2026.100770","url":null,"abstract":"<div><ul><li><span>•</span><span><div>Agentic AI achieved 95.1%-97.2% accuracy in BC SLR screening.</div></span></li><li><span>•</span><span><div>AI annotated 1997 publications in 7.25 h versus 174.7 h for human experts.</div></span></li><li><span>•</span><span><div>Living-OEP reduced annotation time by 95.9% while maintaining accuracy.</div></span></li><li><span>•</span><span><div>Living-OEP demonstrated completeness and guideline concordance in eight scenarios.</div></span></li><li><span>•</span><span><div>Living-OEP is more comprehensive and precise compared with leading chatbots.</div></span></li></ul></div><div><h3>Background</h3><div>Oncologists face increasing difficulty staying current with rapidly evolving clinical data, guidelines, and regulatory updates. Building and maintaining an annotated clinical trial evidence library is time- and labor-intensive. To address this challenge, we developed and validated a living oncology evidence platform (Living-OEP) for breast cancer (BC), powered by an agentic artificial intelligence (AI) system that supports daily, human-conducted, AI-augmented systematic literature review (SLR).</div></div><div>Oncologists face increasing difficulty staying current with rapidly evolving clinical data, guidelines, and regulatory updates. Building and maintaining an annotated clinical trial evidence library is time- and labor-intensive. To address this challenge, we developed and validated a living oncology evidence platform (Living-OEP) for breast cancer (BC), powered by an agentic artificial intelligence (AI) system that supports daily, human-conducted, AI-augmented systematic literature review (SLR).</div><div><h3>Materials and methods</h3><div>The agentic AI system, incorporating GPT-4.1 and o3 (OpenAI) and Claude [Anthropic, PBC, San Francisco, CA] Sonnet-4 (Anthropic), was designed to emulate expert-led, Cochrane-compliant SLR workflows. Guided by a human-developed annotation manual, the system decomposes tasks, self-debugs, and validates outputs. Training data included 29 236 clinical trial abstracts across BC, lung, and prostate cancer, each annotated with four review and 32 extraction variables. Structured data were integrated with guideline-based treatment pathways, forming a real-time, evidence-linked OEP. Accuracy was benchmarked against 1997 human annotations. Living-OEP evidence quality was compared with AI chatbots (ChatGPT [OpenAI, San Francisco, CA], Perplexity [Perplexity AI, Inc., San Francisco, CA], Consensus [Consensus, Boston, MA], and OpenEvidence [OpenEvidence Inc., Miami, FL]) using six criteria across eight BC treatment scenarios.</div></div><div>The agentic AI system, incorporating GPT-4.1 and o3 (OpenAI) and Claude [Anthropic, PBC, San Francisco, CA] Sonnet-4 (Anthropic), was designed to emulate expert-led, Cochrane-compliant SLR workflows. Guided by a human-developed annotation manual, the system decomposes tasks, self-debugs, and validates outputs. Training data included 29 236 clinic","PeriodicalId":100491,"journal":{"name":"ESMO Real World Data and Digital Oncology","volume":"13 ","pages":"Article 100770"},"PeriodicalIF":2.6,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148951075","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}