gms | German Medical Science

72. Jahrestagung der Deutschen Gesellschaft für Neurochirurgie (DGNC)
Joint Meeting mit der Polnischen Gesellschaft für Neurochirurgie

Deutsche Gesellschaft für Neurochirurgie (DGNC) e. V.

06.06. - 09.06.2021

Artikel

Artikel empfehlen

Development and external validation of a clinical prediction model for survival in glioblastoma patients

Die Entwicklung und externe Validierung eines klinischen Prädiktionsmodells für die Überlebenszeiten von Patienten mit Glioblastomen

Meeting Abstract

Suche in Medline nach

  • presenting/speaker Hendrik-Jan Mijderwijk - Universitätsklinikum Düsseldorf, Department of Neurosurgery, Düsseldorf, Deutschland
  • Daan Nieboer - Erasmus MC, Department of Public Health, Rotterdam, Niederlande
  • Fatih Incekara - Erasmus MC, Department of Neurosurgery, Brain Tumor Center, Rotterdam, Deutschland
  • Kerstin Berger - Universitätsklinikum Düsseldorf, Department of Neurosurgery, Düsseldorf, Deutschland
  • Ewout Steyerberg - Erasmus MC, Department of Public Health, Rotterdam, Niederlande; Leiden University Medical Center, Department of Biomedical Data Sciences, Leiden, Niederlande
  • Martin J. van den Bent - Erasmus MC, Department of Neurology, Brain Tumor Center, Rotterdam, Niederlande
  • Guido Reifenberger - Universitätsklinikum Düsseldorf, Department of Neuropathology, Düsseldorf, Deutschland
  • Daniel Hänggi - Universitätsklinikum Düsseldorf, Department of Neurosurgery, Düsseldorf, Deutschland
  • Christian Senft - Universitätsklinikum Jena, Department of Neurosurgery, Jena, Deutschland
  • Marion Rapp - Universitätsklinikum Düsseldorf, Department of Neurosurgery, Düsseldorf, Deutschland
  • Michael Sabel - Universitätsklinikum Düsseldorf, Department of Neurosurgery, Düsseldorf, Deutschland
  • Martin Voß - Universitätsklinikum Frankfurt, Dr. Senckenberg Institute of Neurooncology, Frankfurt am Main, Deutschland
  • Marie-Thérèse Forster - Goethe University Hospital Frankfurt am Main, Department of Neurosurgery, Frankfurt am Main, Deutschland
  • Marcel A. Kamp - Universitätsklinikum Düsseldorf, Department of Neurosurgery, Düsseldorf, Deutschland

Deutsche Gesellschaft für Neurochirurgie. 72. Jahrestagung der Deutschen Gesellschaft für Neurochirurgie (DGNC), Joint Meeting mit der Polnischen Gesellschaft für Neurochirurgie. sine loco [digital], 06.-09.06.2021. Düsseldorf: German Medical Science GMS Publishing House; 2021. DocV077

doi: 10.3205/21dgnc078, urn:nbn:de:0183-21dgnc0781

Veröffentlicht: 4. Juni 2021

© 2021 Mijderwijk et al.
Dieser Artikel ist ein Open-Access-Artikel und steht unter den Lizenzbedingungen der Creative Commons Attribution 4.0 License (Namensnennung). Lizenz-Angaben siehe http://creativecommons.org/licenses/by/4.0/.

Gliederung

Text

Objective: Glioblastoma survival prognostication has become more refined by the molecular reclassification into isocitrate dehydrogenase (IDH) wild-type and IDH mutant tumors. We aimed to provide an updated prediction model that predicts individual survival prognosis in IDH wild-type glioblastoma patients.

Methods: Data from existing databases from Germany and The Netherlands provided data on de novo diagnosed glioblastoma patients treated between 2012 and 2018. The prediction model considered recent glioblastoma biology markers in addition to well-known classical prognostic variables which were updated and refined with additional categories. The clinical prediction model was developed with Cox proportional hazards regression. Performance was evaluated according to calibration (calibration plots, calibration slope) and discrimination (c-statistic) in a cross-validation procedure to assess external validity.

Results: The German patient cohort consisted of 710 patients of whom 511 (72%) had died. Median follow-up was 11 months. The Dutch patient cohort consisted of 326 patients of whom 308 (94.5%) had died. Median follow-up was 10 months. Cohorts (n=1036) were combined to develop three models in order of increasing complexity. The final model considering age, gender, preoperative Karnofsky performance status, extent of surgical resection, MGMT promoter methylation status, and adjuvant therapeutic regime showed an optimism-corrected c-statistic of 0.73 (95% confidence interval 0.71 – 0.75). Cross-validation between the national cohorts yielded comparable results. Moderate miscalibration was observed.

Conclusion: The prediction model reliably predicts individual survival prognosis in newly diagnosed IDH wild-type glioblastoma patients, although additional validation – especially for long-term survival – may be desired. The nomogram and web application support shared decision making.

Figure 1 [Fig. 1]