DP014 | REAL-WORLD DATA ANALYSIS WITH INFORMATICS FOR INTEGRATING BIOLOGY AND THE BEDSIDE (I2B2): A RELIABLE TOOL FOR EVIDENCE GENERATION IN FOLLICULAR LYMPHOMA

Introduction: The increasing accessibility of real-world data (RWD) and advancements in data management tools offer new opportunities for clinical research, addressing limitations of traditional trials. However, challenges such as data quality, availability, and biases persist. This study evaluates the Informatics for Integrating Biology and the Bedside (i2b2) platform for retrospective data collection in follicular lymphoma (FL). Methods: A retrospective study was conducted on adult patients (pts) diagnosed with FL from 2008 to 2019 at ASST Papa Giovanni XXIII, Bergamo, Italy. Clinical data were extracted from electronic health records (EHR) using i2b2, with statistical analyses performed using SAS 9.4. Results: A dedicated extract-transform-load procedure integrated i2b2 with EHR, retrieving structured FL data from multiple sources. Data extraction accuracy was high but varied by data type, availability in structured EHRs, and cross-verification across sources. Text-mining helped recover missing data but with lower accuracy. Critical variables like diagnosis and treatment dates were well captured, but relapse dates required manual revision. A total of 453 pts were included (median age: 63 years; median follow-up: 4.7 years). FLIPI risk distribution was 30.5% low, 30.7% intermediate, and 38.9% high. At diagnosis, 62% received systemic therapy, 10.6% radiotherapy, and 27.4% were managed with watch-and-wait. Among pts treated with radiotherapy, all initially responded, but 27% relapsed. In the watch-and-wait group, 31.5% later required systemic therapy. Overall, 72.2%, 15.9%, and 6.2% of pts required one, two, or three or more lines of systemic therapy, respectively. Disease progression within 24 months occurred in 15.4% of pts, with 58% being high-risk by FLIPI. FLIPI remained prognostic, with high-risk pts experiencing higher relapse rates, greater need for multiple treatments, and increased lymphoma-related mortality. Conclusion: i2b2 is a cost-effective platform for collecting high-quality FL RWD from multiple EHR sources, providing valuable insights into epidemiology and clinical management. The study reaffirms FLIPI’s prognostic value and highlights the need for optimized therapies for high-risk pts. A key limitation was retrieval of data not univocally or consistently present across databases, such as relapse dates. Despite this, i2b2 proves effective in generating real-world evidence, with potential applicability to other diseases.