Resumen:
BCOR and BCORL1 (BCOR/L1) mutations are observed in both myeloid neoplasms (MN) and bone marrow failure syndromes (BMF), but their clinical significance may differ. We analyzed 11 340 patients and identified 418 BCOR/L1-mutated patients, focusing on differences between MN and BMF. Compared with MN, BMF was associated with younger age, a higher frequency of normal karyotype, a greater proportion of missense mutations, and significantly lower variant allele frequencies (VAFs). We next compared BCOR/L1-mutated cases with wild-type cases. In MN, BCOR/L1 mutations were associated with older age, inferior survival, and a higher number of co-occurring mutations, whereas in BMF, they were associated with a higher prevalence of PNH clones. Single-cell DNA sequencing and VAF-based analysis demonstrated that BCOR/L1 mutations were usually secondary in MN but primary events in BMF. Longitudinal sequencing revealed that BCOR/L1-mutant clones often contracted or disappeared over time, even in progression, suggesting limited clonal fitness of BCOR/L1 mutations themselves; instead, progression was associated with increased mutational burden, particularly in RUNX1. Prognostically, BCOR/L1 mutations were associated with inferior survival in MN, whereas in BMF they were not associated with prognosis and tended to correlate with a favorable response to immunosuppressive therapy. RNA-sequencing analyses showed reduced BCOR/L1 expression in BMF, and BCOR knockout was associated with reduced HLA class I expression. Our findings suggest that BCOR/L1 mutations are secondary events in MN and co-occurring mutations may influence disease progression and prognosis, whereas in BMF they appear to be ancestral events acquired independently of progression to MN, potentially through immune privilege-related mechanisms.