Jorne L Biccler1, Sandra Eloranta1, Peter de Nully Brown1, Henrik Frederiksen1, Mats Jerkeman1, Judit Jørgensen1, Lasse Hjort Jakobsen1, Karin E Smedby1, Martin Bøgsted1, Tarec C El-Galaly1. 1. Jorne L. Biccler, Lasse Hjort Jakobsen, Martin Bøgsted, and Tarec C. El-Galaly, Aalborg University Hospital and Aalborg University, Aalborg; Peter de Nully Brown, Copenhagen University Hospital, Copenhagen; Henrik Frederiksen, Odense University Hospital, Odense; Judit Jørgensen, Aarhus University Hospital, Aarhus; Denmark; Sandra Eloranta and Karin E. Smedby, Karolinska Institutet, Stockholm; Mats Jerkeman, Lund University, Lund; and Karin E. Smedby, Karolinska University Hospital, Solna, Sweden.
Abstract
PURPOSE: Prognostic models for diffuse large B-cell lymphoma (DLBCL), such as the International Prognostic Index (IPI) are widely used in clinical practice. The models are typically developed with simplicity in mind and thus do not exploit the full potential of detailed clinical data. This study investigated whether nationwide lymphoma registries containing clinical data and machine learning techniques could prove to be useful for building modern prognostic tools. PATIENTS AND METHODS: This study was based on nationwide lymphoma registries from Denmark and Sweden, which include large amounts of clinicopathologic data. Using the Danish DLBCL cohort, a stacking approach was used to build a new prognostic model that leverages the strengths of different survival models. To compare the performance of the stacking approach with established prognostic models, cross-validation was used to estimate the concordance index (C-index), time-varying area under the curve, and integrated Brier score. Finally, the generalizability was tested by applying the new model to the Swedish cohort. RESULTS: In total, 2,759 and 2,414 patients were included from the Danish and Swedish cohorts, respectively. In the Danish cohort, the stacking approach led to the lowest integrated Brier score, indicating that the survival curves obtained from the stacking model fitted the observed survival the best. The C-index and time-varying area under the curve indicated that the stacked model (C-index: Denmark [DK], 0.756; Sweden [SE], 0.744) had good discriminative capabilities compared with the other considered prognostic models (IPI: DK, 0.662; SE, 0.661; and National Comprehensive Cancer Network-IPI: DK, 0.681; SE, 0.681). Furthermore, these results were reproducible in the independent Swedish cohort. CONCLUSION: A new prognostic model based on machine learning techniques was developed and was shown to significantly outperform established prognostic indices for DLBCL. The model is available at https://lymphomapredictor.org .
PURPOSE: Prognostic models for diffuse large B-cell lymphoma (DLBCL), such as the International Prognostic Index (IPI) are widely used in clinical practice. The models are typically developed with simplicity in mind and thus do not exploit the full potential of detailed clinical data. This study investigated whether nationwide lymphoma registries containing clinical data and machine learning techniques could prove to be useful for building modern prognostic tools. PATIENTS AND METHODS: This study was based on nationwide lymphoma registries from Denmark and Sweden, which include large amounts of clinicopathologic data. Using the Danish DLBCL cohort, a stacking approach was used to build a new prognostic model that leverages the strengths of different survival models. To compare the performance of the stacking approach with established prognostic models, cross-validation was used to estimate the concordance index (C-index), time-varying area under the curve, and integrated Brier score. Finally, the generalizability was tested by applying the new model to the Swedish cohort. RESULTS: In total, 2,759 and 2,414 patients were included from the Danish and Swedish cohorts, respectively. In the Danish cohort, the stacking approach led to the lowest integrated Brier score, indicating that the survival curves obtained from the stacking model fitted the observed survival the best. The C-index and time-varying area under the curve indicated that the stacked model (C-index: Denmark [DK], 0.756; Sweden [SE], 0.744) had good discriminative capabilities compared with the other considered prognostic models (IPI: DK, 0.662; SE, 0.661; and National Comprehensive Cancer Network-IPI: DK, 0.681; SE, 0.681). Furthermore, these results were reproducible in the independent Swedish cohort. CONCLUSION: A new prognostic model based on machine learning techniques was developed and was shown to significantly outperform established prognostic indices for DLBCL. The model is available at https://lymphomapredictor.org .
Authors: R Andrew Harkins; Andres Chang; Sharvil P Patel; Michelle J Lee; Jordan S Goldstein; Selin Merdan; Christopher R Flowers; Jean L Koff Journal: Expert Rev Hematol Date: 2019-09-12 Impact factor: 2.929
Authors: Leila R Zelnick; Michael G Shlipak; Elsayed Z Soliman; Amanda Anderson; Robert Christenson; James Lash; Rajat Deo; Panduranga Rao; Farsad Afshinnia; Jing Chen; Jiang He; Stephen Seliger; Raymond Townsend; Debbie L Cohen; Alan Go; Nisha Bansal Journal: Clin J Am Soc Nephrol Date: 2021-07-12 Impact factor: 10.614
Authors: Adrian Mosquera Orgueira; Marta Sonia González Pérez; Jose Diaz Arias; Laura Rosiñol; Albert Oriol; Ana Isabel Teruel; Joaquin Martinez Lopez; Luis Palomera; Miguel Granell; Maria Jesus Blanchard; Javier de la Rubia; Ana López de la Guia; Rafael Rios; Anna Sureda; Miguel Teodoro Hernandez; Enrique Bengoechea; María José Calasanz; Norma Gutierrez; Maria Luis Martin; Joan Blade; Juan-Jose Lahuerta; Jesús San Miguel; Maria Victoria Mateos Journal: Blood Cancer J Date: 2022-04-25 Impact factor: 9.812
Authors: Selin Merdan; Kritika Subramanian; Turgay Ayer; Johan Van Weyenbergh; Andres Chang; Jean L Koff; Christopher Flowers Journal: Blood Cancer J Date: 2021-01-07 Impact factor: 11.037
Authors: Adrián Mosquera Orgueira; José Ángel Díaz Arias; Miguel Cid López; Andrés Peleteiro Raíndo; Beatriz Antelo Rodríguez; Carlos Aliste Santos; Natalia Alonso Vence; Ángeles Bendaña López; Aitor Abuín Blanco; Laura Bao Pérez; Marta Sonia González Pérez; Manuel Mateo Pérez Encinas; Máximo Francisco Fraga Rodríguez; José Luis Bello López Journal: BMC Cancer Date: 2020-10-21 Impact factor: 4.430