(P 472) LANDSCAPE OF ULTRA-RARE SARCOMAS: A NATIONWIDE STUDY FOR EPIDEMIOLOGY AND PROGNOSIS

Objective: The term "ultra-rare sarcoma" was introduced to highlight the unique clinical difficulties stemming from its extreme rarity. Due to the recency of this classification and the limited volume of existing data, this study sought to explore the epidemiological characteristics and survival outcomes of ultra-rare sarcomas.

Methods: We utilized data from the Bone and Soft Tissue Tumor Registry in Japan (2001–2019) to compare ultra-rare sarcomas with non-ultra-rare sarcomas. Survival outcomes were assessed using Kaplan–Meier estimates with propensity score matching. Additional analyses included multivariate modeling and random survival forest, a machine learning approach.

Results: Out of 22,821 registered cases, 18.9% were diagnosed with ultra-rare sarcomas. Patients with ultra-rare bone sarcomas were significantly older (average age: 57.6 vs. 39.2 years, p < 0.001), whereas those with ultra-rare soft tissue sarcomas were younger (mean age: 49.4 vs. 62.2 years, p < 0.001) than their non-ultra-rare counterparts. Among individuals over 80 years with bone sarcoma and those under 20 with soft tissue sarcoma, nearly half had ultra-rare types. Survival analyses revealed better outcomes for ultra-rare bone sarcoma cases (p = 0.022), but no overall survival difference in soft tissue sarcomas (p = 0.052). However, patients younger than 40 with ultra-rare soft tissue sarcomas exhibited significantly worse survival (p < 0.001). Multivariate analysis yielded hazard ratios of 0.73 for ultra-rare bone sarcomas and 1.25 for ultra-rare soft tissue sarcomas. In the random survival forest model, ultra-rare classification had lower prognostic relevance compared to other variables.

Conclusion: Ultra-rare sarcomas are more frequently observed in older patients with bone sarcomas and younger patients with soft tissue sarcomas. Notably, younger individuals with ultra-rare soft tissue sarcomas face significantly poorer prognoses. Although ultra-rare status plays a relatively minor role in overall prognostic modeling, its impact is age-dependent and clinically relevant in certain subgroups.