Comparing epidemiological and clinical data from RPS patients documented in a German cancer registry to a cohort from TARPSWG reference centres

Purpose: Retroperitoneal sarcomas (RPS) are rare, heterogeneous tumours. Treatment recommendations are mainly derived from cohorts treated at reference centres. The applicability of data from cancer registries (CR) is controversial. This work compares CR and TARPSWG (Transatlantic Australasian Retroperitoneal Sarcoma Working Group) data to assess the representativeness of the TARPSWG and the applicability of the CR data. Methods: TARPSWG cohort has previously been described. The CR Baden-Württemberg cohort includes patients with primary RPS M0 (years 2016–2021, ICD-10 C.49.4/5, C48.x) who underwent surgery within 12 months. Only patients with sarcoma-typical histology codes as used for the German Cancer Society certification system were included. Patient, tumour and therapy factors as well as survival times were compared with Chi2-test, Kaplan Meier curves, and adjusted models. Results: 1000 (TARPSWG) and 364 (CR) patients were included. CR patients were older (median: 64 years vs. 58 years), had more high-grade tumours (FNCLCC 3 48.1% vs. 27.4%, p < 0.0001) and the 5-year survival rate was significantly lower (56.3% vs. 67.9%, p = 0.0015). The proportions of dedifferentiated liposarcoma (CR 37.1% vs. 37.0%) and leiomyosarcoma (CR 20.1% vs. 19.2%), and patterns of recurrence in these most frequent RPS subtypes were similar. Conclusion: ICD-O/ICD 10 based filters appear to be a valid tool for extracting RPS cases from CR. The similar distribution and biological behavior of distinct RPS subtypes suggests that TARPS-WG are representative, and CR data may be used to verify recommendations derived from reference centre cohorts. Complementary use of data from different sources warrants further investigation in rare cancers.