Abstract 77P
Background
The European Reference Network on rare adult solid cancers (EURACAN), aims to develop a European registry for sarcomas leveraging databases already available. Due to the high number of centres involved, the harmonization of the data and management of possible heterogeneity of data between centres is complex. We aimed to test whether the Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM) was a possible solution to address data heterogeneity.
Methods
Three registries converted their data to the OMOP-CDM: the population-based Netherlands Cancer Registry (NCR) and two sarcomas clinical registries (Istituto Nazionale dei Tumori, INT and University Hospital Graz, Graz). Using retroperitoneal sarcoma (RPS) as a use case, we analysed data mapped to the OMOP CDM to assess prognostic factors in RPS. We selected patients ≥18 years old with a primary localized RPS with a surgery between 01/01/2010 and 31/12/2017. We compared patient, tumour and treatment characteristics across registries. We computed 5-year overall survival (OS) by sex and histology grouping to compare results with available evidence. Analyses to identify prognostic factors are ongoing.
Results
We identified 848 patients with RPS. Patient characteristics were similar across registries. The most common histologies were liposarcomas and leiomyosarcomas. INT had a higher percentage of dedifferentiated liposarcomas (43%) compared to Graz (37%) and NCR (27%). High-grade (G2+G3) RPS were very common (over 60%) in clinical settings (INT, Graz). Grading in NCR is not comparable. Perioperative chemotherapy was used in 26% of RPS at INT and it was rarely used in Graz. Perioperative radiotherapy was used in 18%, 19% and 47% of RPS at INT, NCR and Graz, respectively. Surgery was macroscopically complete (R0/R1) in almost all cases of INT and Graz. OS was 78% (95% CI: 60%-100%), 74% (95% CI %: 70%-79%) and 65% (95% CI: 60%-69%) in Graz, INT and NCR respectively.
Conclusions
This use case demonstrated that the OMOP CDM represents a reliable oncology model. The findings are consistent with expert knowledge and available evidence. It is worth highlighting that comparison between different contexts may be affected by changes in sarcoma classification over time.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
Fondazione IRCCS Istituto Nazionale dei Tumori di Milano.
Funding
Has not received any funding.
Disclosure
All authors have declared no conflicts of interest.