Dataset Information

Personalized risk prediction for event-free survival at 24 months in patients with diffuse large B-cell lymphoma.

ABSTRACT: We recently defined event-free survival at 24 months (EFS24) as a clinically relevant outcome for patients with DLBCL. Patients who fail EFS24 have very poor overall survival, while those who achieve EFS24 have a subsequent overall survival equivalent to that of the age- and sex-matched general population. Here, we develop and validate a clinical risk calculator (IPI24) for EFS24. Model building was performed on a discovery dataset of 1,348 patients with DLBCL and treated with anthracycline-based immunochemotherapy. A multivariable model containing age, Ann Arbor stage, normalized serum LDH, ALC, ECOG performance status, bulky disease, and sex was identified. The model was then applied to an independent validation dataset of 1,177 DLBCL patients. The IPI24 score estimates the probability of failing to achieve the EFS24 endpoint for an individual patient. The IPI24 model showed superior discriminatory ability (c-statistic?=?0.671) in the validation dataset compared to the IPI (c-statistic?=?0.649) or the NCCN-IPI (c-statistic?=?0.657). After recalibration of the model on the combined dataset, the median predicted probability of failing to achieve EFS24 was 36% (range, 12-88%), and the IPI24 showed an EFS24 gradient in all IPI groups. The IPI24 also identified a significant percentage of patients with high risk disease, with over 20% of patients having a 50% or higher risk of failing to achieve EFS24. The IPI24 provides an individual patient level probability of achieving the clinically relevant EFS24 endpoint. It can be used via electronic apps.

SUBMITTER: Maurer MJ

PROVIDER: S-EPMC4801345 | biostudies-literature | 2016 Feb

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Personalized risk prediction for event-free survival at 24 months in patients with diffuse large B-cell lymphoma.

Maurer Matthew J MJ Jais Jean-Philippe JP Ghesquières Hervé H Witzig Thomas E TE Hong Fangxin F Haioun Corinne C Thompson Carrie A CA Thieblemont Catherine C Micallef Ivana N IN Porrata Luis F LF Ribrag Vincent V Nowakowski Gregorz S GS Casasnovas Olivier O Bologna Serge S Morschhauser Franck F Morrison Vicki A VA Peterson Bruce A BA Macon William R WR Copie-Bergman Christiane C Feldman Andrew L AL Syrbu Sergei I SI Kurtin Paul J PJ Gascoyne Randy D RD Li Hailun H Allmer Cristine C Kahl Brad S BS Ansell Stephen M SM Slager Susan L SL Link Brian K BK Salles Gilles G Habermann Thomas M TM Tilly Hervé H Cerhan James R JR

American journal of hematology 20151126 2

We recently defined event-free survival at 24 months (EFS24) as a clinically relevant outcome for patients with DLBCL. Patients who fail EFS24 have very poor overall survival, while those who achieve EFS24 have a subsequent overall survival equivalent to that of the age- and sex-matched general population. Here, we develop and validate a clinical risk calculator (IPI24) for EFS24. Model building was performed on a discovery dataset of 1,348 patients with DLBCL and treated with anthracycline-base ...[more]

PMID: 26492520

Similar Datasets

Project description:BackgroundThirty to forty percent of patients with Diffuse Large B-cell Lymphoma (DLBCL) have an adverse clinical evolution. The increased understanding of DLBCL biology has shed light on the clinical evolution of this pathology, leading to the discovery of prognostic factors based on gene expression data, genomic rearrangements and mutational subgroups. Nevertheless, additional efforts are needed in order to enable survival predictions at the patient level. In this study we investigated new machine learning-based models of survival using transcriptomic and clinical data.MethodsGene expression profiling (GEP) of in 2 different publicly available retrospective DLBCL cohorts were analyzed. Cox regression and unsupervised clustering were performed in order to identify probes associated with overall survival on the largest cohort. Random forests were created to model survival using combinations of GEP data, COO classification and clinical information. Cross-validation was used to compare model results in the training set, and Harrel's concordance index (c-index) was used to assess model's predictability. Results were validated in an independent test set.ResultsTwo hundred thirty-three and sixty-four patients were included in the training and test set, respectively. Initially we derived and validated a 4-gene expression clusterization that was independently associated with lower survival in 20% of patients. This pattern included the following genes: TNFRSF9, BIRC3, BCL2L1 and G3BP2. Thereafter, we applied machine-learning models to predict survival. A set of 102 genes was highly predictive of disease outcome, outperforming available clinical information and COO classification. The final best model integrated clinical information, COO classification, 4-gene-based clusterization and the expression levels of 50 individual genes (training set c-index, 0.8404, test set c-index, 0.7942).ConclusionOur results indicate that DLBCL survival models based on the application of machine learning algorithms to gene expression and clinical data can largely outperform other important prognostic variables such as disease stage and COO. Head-to-head comparisons with other risk stratification models are needed to compare its usefulness.

Project description:We performed a multistage genome-wide association study to identify inherited genetic variants that predict outcome in diffuse large B-cell lymphoma patients treated with immunochemotherapy.We conducted a meta-analysis of two genome-wide association study data sets, one from the LNH2003B trial (N = 540), a prospective clinical trial from the Lymphoma Study Association, and the other from the Molecular Epidemiology Resource study (N = 312), a prospective observational study from the University of Iowa-Mayo Clinic Lymphoma Specialized Program of Research Excellence. Top single nucleotide polymorphisms were then genotyped in independent cohorts of patients from the Specialized Program of Research Excellence (N = 391) and the Groupe Ouest-Est des Leucémies Aiguës et Maladies du Sang (GOELAMS) -075 randomized trial (N = 294). We calculated the hazard ratios (HRs) and 95% CIs for event-free survival (EFS) and overall survival (OS) using a log-additive genetic model with adjustment for age, sex, and age-adjusted International Prognostic Index.In a meta-analysis of the four studies, the top loci for EFS were marked by rs7712513 at 5q23.2 (near SNX2 and SNCAIP; HR, 1.39; 95% CI, 1.23 to 1.57; P = 2.08 × 10(-7)), and rs7765004 at 6q21 (near MARCKS and HDAC2; HR, 1.38; 95% CI, 1.22 to 1.57; P = 7.09 × 10(-7)), although they did not reach conventional genome-wide significance (P = 5 × 10(-8)). Both rs7712513 (HR, 1.49; 95% CI, 1.29 to 1.72; P = 3.53 × 10(-8)) and rs7765004 (HR, 1.47; 95% CI, 1.27 to 1.71; P = 5.36 × 10(-7)) were also associated with OS. In exploratory analyses, a two-single nucleotide polymorphism risk score was highly predictive of EFS (P = 1.78 × 10(-12)) and was independent of treatment, IPI, and cell-of-origin classification.Our study provides encouraging evidence for associations between loci at 5q23.2 and 6q21 with EFS and OS in patients with diffuse large B-cell lymphoma treated with immunochemotherapy, suggesting novel biology and the potential contribution of host genetics to the prognosis of this aggressive malignancy.

Dataset Information

Personalized risk prediction for event-free survival at 24 months in patients with diffuse large B-cell lymphoma.

Publications

Personalized risk prediction for event-free survival at 24 months in patients with diffuse large B-cell lymphoma.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure