scholarly journals Synthetizing Published Evidence on Survival by Reconstruction of Patient-Level Data and Generation of a Multi-Trial Kaplan-Meier Curve

Cureus ◽  
2021 ◽  
Author(s):  
Andrea Messori
Keyword(s):  
Patient Level Data ◽  
Patient Level ◽  
Kaplan Meier ◽  
Published Evidence ◽  
Level Data

scholarly journals Efficacy of Ibrutinib-Rituximab versus Real-World (RW) Treatments for Patients with Waldenström's Macroglobulinemia (WM): Adjusted Comparison of iNNOVATE and the Lyon-Sud RW Database

Blood ◽  
2018 ◽  
Vol 132 (Supplement 1) ◽  
pp. 1604-1604
Author(s):  
Lionel Karlin ◽  
Hervé Besson ◽  
Christoph Tapprich ◽  
Jamie Garside ◽  
Gilles Salles
Keyword(s):  
Clinical Practice ◽  
Board Of Directors ◽  
Real World ◽  
Research Funding ◽  
Advisory Committees ◽  
Patient Level Data ◽  
Treatment Regimens ◽  
Patient Level ◽  
Kaplan Meier ◽  
Level Data

Abstract Introduction Ibrutinib, a first-in-class, oral, covalent inhibitor of Bruton's tyrosine kinase, is approved in several countries as monotherapy for adults with WM, based on the high response rate (90.5%) observed in a phase 2, single-arm trial of symptomatic patients who had received ≥ 1 prior therapy (NCT01614821; Treon SP, et al. N Engl J Med. 2015;372:1430-1440). In the phase 3 iNNOVATE trial (NCT02165397; Dimopoulos MA, et al. N Engl J Med. 2018;378:2399-2410), adding rituximab to ibrutinib (IR) led to a statistically significant improvement in progression-free survival (PFS) compared with placebo-rituximab, in both treatment-naïve (TN) patients and those with disease recurrence. This analysis examined the relative treatment effect of IR versus other RW treatment regimens used in daily clinical practice in TN and relapsed/refractory (R/R) WM. An adjusted comparison was conducted using patient-level data from the iNNOVATE trial and the Lyon-Sud RW database in France. Methods The Lyon-Sud database holds medical records for patients with WM diagnosed between 1980 and 2017 from the Centre Hospitalier Lyon-Sud. PFS and overall survival (OS) outcomes were compared between the IR arm of iNNOVATE and RW physicians' choice (PC) treatment in Lyon-Sud (excluding RW ibrutinib). Kaplan-Meier survival curves by patient cohort were generated for both end points. A multivariate Cox proportional hazards model was fitted on the pooled patient-level data from both sources to estimate adjusted hazard ratios (HRs) for effect of IR on PFS and OS versus RW treatment, with age, sex, and treatment line as covariates. The unit of observation for the RW databases was the treatment line (rather than the patient) number, whereby RW patients receiving multiple treatment lines contributed to multiple observations, and baseline was defined as the line-specific treatment start date. Results Overall, 242 treatment lines from patients with WM were identified from the Lyon-Sud database; 224 were analyzed. Baseline characteristics were comparable between the IR arm of iNNOVATE (n = 75) and the RW cohort: 62.7% versus 66.5% ≥ 65 years of age; line of therapy, 45.3% versus 48.7% TN [1L], 25.3% versus 25.0% second-line [2L], and 29.3% versus 26.3% third-line or later [3L+]. Median follow-up was 26.7 months for iNNOVATE and 68.5 months for the RW cohort. Non-ibrutinib PC regimens in the RW cohort included rituximab (n = 51 treatment lines), chemotherapy (n = 66; including chlorambucil, n = 31), rituximab-cyclophosphamide-dexamethasone (n = 35), rituximab-CHOP/CHOP-like (n = 21), rituximab-chlorambucil (n = 15), fludarabine-cyclophosphamide-rituximab (n = 14), bendamustine-rituximab (n = 10), other rituximab-containing chemoimmunotherapy (n = 10), and rituximab-targeted agent (n = 2). Two ibrutinib lines were excluded from the RW cohort analysis. Figures A and B show the observed 1L+ Kaplan-Meier curves for IR versus RW PC therapy for all analyzed patients with WM (unadjusted HRs: 0.32 [95% confidence interval (CI), 0.19-0.55] for PFS and 0.38 [95% CI, 0.13-1.09] for OS). After adjusting for differences in patient characteristics, HRs (1L+ therapy) became 0.28 (95% CI, 0.16-0.48; p < 0.001) for PFS and 0.29 (95% CI, 0.09-0.93; p = 0.037) for OS. Restricting the analysis to 1L treatment only (Lyon-Sud n = 109; iNNOVATE n = 34), the adjusted HRs were 0.25 (95% CI, 0.09-0.70; p < 0.009) for PFS and 0.20 (95% CI, 0.02-2.00; p = 0.170) for OS; respective unadjusted HRs were 0.31 (95% CI, 0.12-0.78) and 0.30 (95% CI, 0.04-2.34). In the 2L+ setting (Lyon-Sud n = 115; iNNOVATE n = 41), the adjusted HRs were 0.28 (95% CI, 0.15-0.56; p < 0.001) for PFS and 0.34 (95% CI, 0.08-1.35; p = 0.126) for OS; respective unadjusted HRs were 0.31 (95% CI, 0.16-0.61) and 0.39 (95% CI, 0.12-1.31). Concl usions In the absence of randomized controlled trial data for ibrutinib versus other treatment regimens for WM, these adjusted comparisons of clinical trial and RW patient-level data suggest that IR significantly improves both PFS and OS versus RW PC regimens as 1L+ therapy. These results help inform physicians on the standard of care for WM in clinical practice. Funding Source: This project was sponsored by Janssen Pharmaceutica NV, and Pharmacyclics LLC, an AbbVie Company. The real-world databases are independently owned. Writing assistance was provided by Emma Fulkes of PAREXEL and funded by Janssen Pharmaceutica NV. Disclosures Karlin: Celgene: Honoraria, Membership on an entity's Board of Directors or advisory committees; Janssen: Honoraria, Membership on an entity's Board of Directors or advisory committees, Other: travel support; Amgen: Honoraria, Membership on an entity's Board of Directors or advisory committees, Other: travel support. Besson:Janssen Pharmaceutica NV: Employment. Tapprich:Janssen Pharmaceutica NV: Employment. Garside:Janssen Pharmaceutica NV: Employment. Salles:Roche: Honoraria, Research Funding; Celgene: Honoraria, Research Funding; AbbVie: Honoraria; Acerta: Honoraria; Amgen: Honoraria; Novartis: Consultancy, Honoraria; Merck: Honoraria; Pfizer: Honoraria; Takeda: Honoraria; Servier: Honoraria; Janssen: Honoraria; Gilead: Honoraria; Epizyme: Honoraria; Morphosys: Honoraria.

scholarly journals Amalgamation of cloud-based colonoscopy videos with patient-level metadata to facilitate large-scale machine learning

2021 ◽  
Vol 09 (02) ◽  
pp. E233-E238
Author(s):  
Rajesh N. Keswani ◽  
Daniel Byrd ◽  
Florencia Garcia Vicente ◽  
J. Alex Heller ◽  
Matthew Klug ◽  
...  
Keyword(s):  
Machine Learning ◽  
Large Scale ◽  
Medical Center ◽  
Academic Medical Center ◽  
Procedure Time ◽  
Image Capture ◽  
Common Procedure ◽  
Patient Level Data ◽  
Patient Level ◽  
Level Data

Abstract Background and study aims Storage of full-length endoscopic procedures is becoming increasingly popular. To facilitate large-scale machine learning (ML) focused on clinical outcomes, these videos must be merged with the patient-level data in the electronic health record (EHR). Our aim was to present a method of accurately linking patient-level EHR data with cloud stored colonoscopy videos. Methods This study was conducted at a single academic medical center. Most procedure videos are automatically uploaded to the cloud server but are identified only by procedure time and procedure room. We developed and then tested an algorithm to match recorded videos with corresponding exams in the EHR based upon procedure time and room and subsequently extract frames of interest. Results Among 28,611 total colonoscopies performed over the study period, 21,170 colonoscopy videos in 20,420 unique patients (54.2 % male, median age 58) were matched to EHR data. Of 100 randomly sampled videos, appropriate matching was manually confirmed in all. In total, these videos represented 489,721 minutes of colonoscopy performed by 50 endoscopists (median 214 colonoscopies per endoscopist). The most common procedure indications were polyp screening (47.3 %), surveillance (28.9 %) and inflammatory bowel disease (9.4 %). From these videos, we extracted procedure highlights (identified by image capture; mean 8.5 per colonoscopy) and surrounding frames. Conclusions We report the successful merging of a large database of endoscopy videos stored with limited identifiers to rich patient-level data in a highly accurate manner. This technique facilitates the development of ML algorithms based upon relevant patient outcomes.

scholarly journals Clinical knowledge extraction via sparse embedding regression (KESER) with multi-center large scale electronic health record data

2021 ◽  
Vol 4 (1) ◽  
Author(s):  
Chuan Hong ◽  
Everett Rush ◽  
Molei Liu ◽  
Doudou Zhou ◽  
Jiehuan Sun ◽  
...  
Keyword(s):  
Feature Selection ◽  
Electronic Health Record ◽  
Large Scale ◽  
Knowledge Extraction ◽  
Knowledge Map ◽  
Health Record ◽  
Clinical Knowledge ◽  
Patient Level Data ◽  
Patient Level ◽  
Level Data

AbstractThe increasing availability of electronic health record (EHR) systems has created enormous potential for translational research. However, it is difficult to know all the relevant codes related to a phenotype due to the large number of codes available. Traditional data mining approaches often require the use of patient-level data, which hinders the ability to share data across institutions. In this project, we demonstrate that multi-center large-scale code embeddings can be used to efficiently identify relevant features related to a disease of interest. We constructed large-scale code embeddings for a wide range of codified concepts from EHRs from two large medical centers. We developed knowledge extraction via sparse embedding regression (KESER) for feature selection and integrative network analysis. We evaluated the quality of the code embeddings and assessed the performance of KESER in feature selection for eight diseases. Besides, we developed an integrated clinical knowledge map combining embedding data from both institutions. The features selected by KESER were comprehensive compared to lists of codified data generated by domain experts. Features identified via KESER resulted in comparable performance to those built upon features selected manually or with patient-level data. The knowledge map created using an integrative analysis identified disease-disease and disease-drug pairs more accurately compared to those identified using single institution data. Analysis of code embeddings via KESER can effectively reveal clinical knowledge and infer relatedness among codified concepts. KESER bypasses the need for patient-level data in individual analyses providing a significant advance in enabling multi-center studies using EHR data.

Sign in / Sign up

Export Citation Format

Share Document